docs(changelog): v2.6.1-phase1-opencode

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

.gitignore

@@ -16,5 +16,5 @@ data/*
 !data/AGENTS.md
 !data/skills/
 !data/mcp.json
-!data/coder-providers.json
+!data/coder-providers.example.json
 codecontext/fork.tar.gz

BOOCODER.md

@@ -37,3 +37,81 @@ Every file modification queues in `pending_changes` before touching disk. The us
 - Never count `dist/` directory sizes as source lines. Only count `src/**/*.ts` files. Compiled output is inflated by inlined types and transpilation artifacts.
 - Before claiming a feature works, run the actual command and show the output. "Should work" is not verification. Acceptable evidence: test output (`pnpm test`), build output (`pnpm build`), curl response, docker logs, `\d tablename` output. If you can't run it, say so explicitly — don't assert success without evidence.
 - When reporting counts (tools, tests, files, routes, lines), derive the number from a command (`grep -c`, `wc -l`, test runner output) — not from memory or approximation.
+
+## Provider lifecycle (v2.3)
+
+BooCoder's coding agents are a **config-backed registry**: built-ins live in `provider-registry.ts`, and `data/coder-providers.json` layers overrides + custom entries on top. Registration ≠ installation — the config lists what you *want*; a probe reports what's *ready*.
+
+### Config file: `data/coder-providers.json`
+
+Resolved from `CODER_PROVIDERS_PATH` (default `/data/coder-providers.json`; dev/host path `/opt/boocode/data/coder-providers.json`). It is **gitignored** — it's live runtime config that the coder reads *and writes* (UI toggles `PATCH` it), so tracking it would churn `git status`. The tracked reference is `data/coder-providers.example.json`; copy it to `coder-providers.json` to seed overrides. A missing file, invalid JSON, or a schema mismatch all fall back to built-ins-only — loading never throws at startup.
+
+```json
+{
+  "providers": {
+    "goose": { "enabled": false },
+    "amp-acp": {
+      "extends": "acp",
+      "label": "Amp",
+      "description": "ACP wrapper for Amp",
+      "command": ["amp-acp"],
+      "enabled": true
+    }
+  }
+}
+```
+
+Per-provider override fields (all optional):
+
+| Field | Meaning |
+|-------|---------|
+| `extends` | `"acp"` — required for a NEW (custom) provider; built-in overrides omit it |
+| `label` | Display name (required for custom) |
+| `description` | Sub-label shown in the picker / settings |
+| `command` | `[binary, ...args]` to spawn (required for custom; overrides a built-in's default argv) |
+| `env` | Extra env vars merged into the spawn |
+| `enabled` | Default `true`; `false` hides it from the composer |
+| `order` | UI sort key |
+| `models` / `additionalModels` | Replace / merge onto the discovered model list |
+
+A PATCH to one provider id **replaces that id's override object wholesale** (per-id shallow merge), so to flip a single field keep the rest; a `null` value for an id deletes its override (reverts to the built-in default).
+
+### Refresh contract
+
+The snapshot is cached and a provider's cold ACP probe (tier-2) is **skipped** while `available_agents.last_probed_at` is younger than `PROVIDER_PROBE_TTL_MS` (default `86400000` = 24h). Opening the composer is therefore fast and does not re-probe. To force a cold re-probe (after installing a CLI or editing models): **`POST /api/providers/refresh`** (the Refresh button in the Providers settings tab), which clears the cache and re-probes.
+
+### Enable / disable
+
+Two ways:
+- **Settings → Providers tab** — open the sidebar → **Settings** → **Providers**: toggle a provider on/off, refresh it, or open its diagnostic. (Earlier builds exposed a gear in the composer; that control was moved into Settings.)
+- **Edit the config** (`"enabled": false`) then `POST /api/providers/refresh`.
+
+A **disabled** provider leaves the composer's provider picker but stays listed in the Providers tab (status "Disabled") so you can re-enable it. **Native `boocode` is always-on** — an `enabled:false` on it is ignored (with a warn log) and it is never rendered as toggleable.
+
+### Adding a custom ACP provider
+
+- **Catalog modal**: Providers tab → **Add provider** → pick an entry → it PATCHes the config (`extends:'acp'` + label + command, enabled) and refreshes that provider.
+- **Hand-edit** `data/coder-providers.json`: add an id with `extends:'acp'`, `label`, and `command`, then `POST /api/providers/refresh`.
+
+Either way, **adding to config does NOT install the binary.** Until the CLI is on `PATH` the provider shows **"Not installed"** (status `unavailable`) and does not appear in the composer picker.
+
+### Known limitation — subset refresh
+
+`POST /api/providers/refresh` accepts an optional `{ "providers": ["id", ...] }` body and returns a `refreshed` count scoped to that subset — **but the underlying cold re-probe currently covers ALL installed providers**, not just the requested subset. True per-provider force is a future change (it needs a snapshot-internal parameter). This is intentional for now, not a bug: a subset refresh still re-probes everything; only the reported count is scoped.
+
+### Deploy + smoke
+
+Two deploy targets:
+- **Routes (host service):** `pnpm -C apps/server build && pnpm -C apps/coder build && sudo systemctl restart boocoder`
+- **Web UI (container):** `docker compose up --build -d boocode`
+
+Green gate (verified across phases 1–5): `pnpm -C apps/coder test` (134 passing) `&& pnpm -C apps/coder build`.
+
+Smoke (via Tailscale):
+
+```bash
+curl http://100.114.205.53:9502/api/providers/snapshot       # lists every registered provider
+curl http://100.114.205.53:9500/api/coder/providers/config   # raw config, through the BooChat proxy
+# Settings → Providers: disable goose → it leaves the composer picker, stays in the tab
+# POST refresh → models repopulate; Add a catalog entry → it appears after refresh (unavailable until its CLI is installed)
+```

CHANGELOG.md

@@ -2,6 +2,46 @@
 
 All notable changes per release tag. Most recent on top, ordered by tag creation date (which matches the git history). Tag names follow `vMAJOR.MINOR.PATCH-slug` — the slug describes what shipped, so the tag name alone is enough to recall the batch.
 
+## v2.6.1-phase1-opencode — 2026-05-30
+
+v2.6 Phase 1: opencode runs as a warm HTTP server (`apps/coder/src/services/backends/opencode-server.ts`) — one `opencode serve` per BooCoder process, one opencode session per BooCode session resumed across turns via the new `agent_sessions` table, with a single SSE read loop, reasoning dedup ported from Paseo, an inactivity watchdog, and a stale-session guard (crashed-not-resumed + a `config_hash` fingerprint over `opencode_server|<model>`, deliberately excluding the ephemeral server port so cross-restart resume survives). Builds on the `v2.6.0-phase0-foundations` schema/interface scaffold. The batch's hard-won fixes: opencode streams `session.next.*` events (not `message.part.*`), and `event.subscribe()` must pass the session's worktree `directory` or events route to the server CWD and turns come back empty; model strings must be `llama-swap/`-prefixed and present in opencode's own config, with `agent-probe` now populating `available_agents.models` via `mergeLlamaSwap` so the frontend stops sending an empty model; `session_worktrees`/`agent_sessions` FKs are `ON DELETE CASCADE` so session deletion no longer 500s. Also bundled: dcp-message-id tag stripping from opencode text output, a reopen-closed-pane control, the `[+]`/split-pane button separation, auto-name using the session's loaded model, and a `systematic-debugging` slash command. Smoke 1 verified end-to-end (two turns, session reuse, turn 2 ~9x faster). Known Phase 1 limit: one SSE stream scoped to the most-recent session's directory — concurrent opencode sessions in different worktrees collide (warns; per-session SSE is Phase 2).
+
+## v2.5.15-acp-path-guard — 2026-05-29
+
+Security fix + repo hygiene. Fixes a path-traversal in the ACP filesystem bridge (`acp-client-fs.ts`, flagged by the automated push security review): the worktree guard used an unbounded `startsWith(resolve(worktreePath))`, so a sibling path sharing the worktree as a string prefix (`<worktree>-evil/…`) escaped the scope — and `writeWorktreeTextFile` writes to disk directly (no `pending_changes` gate), so a confused/buggy ACP agent could write outside its worktree. Now uses a separator-bounded check matching `write_guard.ts` (`resolve()` + `startsWith(root + sep)` / `=== root`) via a shared `resolveInWorktree`, with a regression test covering `../` traversal and the sibling-prefix bug. Symlink-swap/`O_NOFOLLOW` hardening was intentionally skipped — consistent with `write_guard`'s no-realpath stance, and the agent already runs with host FS access so this is a containment guard, not a trust boundary. Separately, stops tracking the live `data/coder-providers.json` (it's runtime config the UI reads *and writes* on provider toggles, which churned `git status`) — it's now gitignored with a tracked `data/coder-providers.example.json` reference; the loader falls back to built-ins-only when the live file is absent. The provider-type duplication (coder ↔ web) stays guarded by the existing text-identity `provider-types-parity.test.ts` — a shared package was considered and declined (drift is already prevented; not worth the Docker/build-order risk at solo scale).
+
+## v2.5.14-claude-md — 2026-05-29
+
+Docs-only — CLAUDE.md session-learnings update, no code. Adds gotchas surfaced while shipping the v2.3 provider-lifecycle batch: the host `boocoder.service` keeps running the old process after `pnpm -C apps/coder build` (stale-process tell = new routes 404 while old routes 200, restart don't re-debug); the `boocode` container `build: .` deploys the working tree, so web edits are live on the Vite dev server but not production until `docker compose up --build -d boocode`; `PATCH /api/providers/config` replaces a provider's override wholesale (send `{...existing, enabled}` or a custom ACP entry's command is wiped) and `data/coder-providers.json` is live config not to be committed as code; external agents dispatch one-shot with no context/token tracking (only native `boocode` tracks ctx; OpenCode-as-server is the unshipped `v2-6-persistent-agent-sessions` plan); the `ui/` primitive inventory with `button role=switch` / Dialog fallbacks for the absent switch/sheet; and the mobile Dialog-with-list scroll-containment recipe. Also backfills previously-uncommitted doc bullets for the `v2.5.7`–`v2.5.11` coder work (provider-type parity test, async ACP command discovery, AgentComposerBar `installed` filter, provider-registry path disambiguation).
+
+## v2.5.13-provider-lifecycle-phase5 — 2026-05-29
+
+Closeout of the v2.3 provider-lifecycle batch — the web UI (Phase 5) plus docs (Phase 6). Provider management moved into **Settings → Providers**: a tab listing every registered provider with a status badge (Available / Disabled / Not installed / Error / Loading), an enable/disable toggle, a per-provider refresh, and a plaintext diagnostic; toggling sends the provider's *full* override (preserving a custom ACP entry's command under the wholesale-replace PATCH merge) then refetches the snapshot. The composer's provider picker now filters to `enabled && (status === 'ready' || 'loading')`, so disabled and unavailable providers drop out of the picker and are managed only in settings (native `boocode` always shows). A curated ACP catalog (`apps/web/src/data/acp-provider-catalog.ts`) + `AddProviderModal` register custom providers via `PATCH /api/providers/config` then a subset refresh, and the web client gained `getProvidersConfig` / `patchProvidersConfig` / `refreshProviders` / `getProviderDiagnostic`. Two mobile fixes ship alongside: the Settings pane is now reachable on phones (opening it pushes `?pane=` atomically so the mobile URL-sync effect keeps it active instead of snapping back to the chat pane), and the Add-provider modal caps to the viewport with a single `overscroll-contain` scroll region so the list scrolls instead of dragging the whole modal. This completes the arc begun in `v2.5.4-provider-lifecycle-phase1` (config-backed registry over the built-ins) → `v2.5.5-provider-lifecycle-phase2` (loading/unavailable snapshot lifecycle + tier-2 probe TTL gate) → `v2.5.6-provider-lifecycle-phase3` (generic `resolveLaunchSpec` ACP dispatch) → `v2.5.12-provider-lifecycle-phase4` (config GET/PATCH, subset refresh, diagnostic HTTP API). Docs landed in `BOOCODER.md` (config file, refresh contract, enable/disable, custom ACP, the honest subset-refresh known limitation) and `docs/DEFERRED-WORK.md` §2 is marked addressed; the remaining Tier-2 follow-ups (WS `provider_snapshot_updated` frame, `available_agents.enabled` column, shared types package, MCP provider tools) stay deferred.
+
+## v2.5.12-provider-lifecycle-phase4 — 2026-05-29
+
+Phase 4 of the v2.3 provider-lifecycle batch (`openspec/changes/v2-3-provider-lifecycle/design.md` §6): the HTTP API to read, patch, refresh, and diagnose providers. `routes/providers.ts` gains `GET /api/providers/config` (the raw loaded `CoderProvidersFile`), `PATCH /api/providers/config` (a partial providers map — an id's override object is replaced wholesale, a `null` value deletes it), an optional `{ providers?: string[] }` body on `POST /api/providers/refresh` (the `refreshed` count reflects the requested subset; the force probe itself still covers all installed providers, since per-provider force is a snapshot-internal change left to a later phase), and `GET /api/providers/:id/diagnostic` returning JSON `{ diagnostic: string }` — a read-only report (resolved def, install_path, last_probed_at, enabled, `which` availability, last cached probe error) with no probe spawn. PATCH correctness is the whole story: the order is validate→save→reload→clear, a malformed body or an invalid merged config returns 422 without writing the file, and a `save()` failure returns 500 without reloading the registry or clearing the snapshot cache, so on-disk and in-memory state can never diverge. New pure `mergeProviderConfigPatch` + `ProviderConfigPatchSchema` in `provider-config.ts`, a read-only `peekSnapshotEntry` cache accessor (source of the diagnostic's last-error — no probe/cache logic change), and a new `provider-diagnostic.ts` formatter. The web client gains `api.coder.getProvidersConfig` / `patchProvidersConfig` / `refreshProviders(providers?)` / `getProviderDiagnostic`, with mirrored `ProviderOverride` / `CoderProvidersFile` / `ProviderConfigPatch` types; the existing `/api/coder/*` proxy blanket-forwards the new routes with no change. +28 tests (134 coder total: pure merge/validate, the diagnostic formatter, and `app.inject` route tests proving the 422-no-write and save-fail-no-divergence guards). The diagnostic returns JSON rather than the §8 plaintext so it flows through the JSON `request` client helper (reconciling design §6.4's `{ diagnostic }` with §8's string report). No UI (Phase 5). Builds on `v2.5.6-provider-lifecycle-phase3`.
+
+## v2.5.11-claude-skill-discovery — 2026-05-29
+
+Surface Claude Code's real enabled commands + plugin skills in the coder slash menu, with icons separating commands from plugin skills. New `claude-command-discovery.ts` reads (user-global scope) `~/.claude/commands/*.md` plus every enabled plugin in `~/.claude/settings.json:enabledPlugins` — each plugin's user-scope install path contributes `skills/<name>/SKILL.md` (kind `skill`) and `commands/*.md` (kind `command`), parsed from frontmatter, bare names, deduped. The snapshot's claude branch discovers these **live** (claude is PTY, no ACP probe; the snapshot cache rate-limits the fs reads). The `/` menu now renders up to three icon'd groups: **`<agent> commands`** (Terminal), **`<agent> skills`** (Puzzle — claude's plugin skills / opencode is all commands), and **BooCoder skills** (Sparkles), via a new optional `icon` on `SlashCommandGroup`. `AgentCommand` gains a `kind` field, added identically to the coder and web copies (the `provider-types-parity` test enforces it); `mergeCommandsByName` is now generic so it preserves the tag. Invocation is unchanged — picking a claude command/skill sends `/name` to claude (PTY), which executes it. Project-local plugins + `<cwd>/.claude/commands` deferred. BooChat unaffected (flat skills). Smoke-test the claude skill slash-execution on the host.
+
+## v2.5.10-opencode-live-commands — 2026-05-29
+
+Surface opencode's real (live ACP) command set in the coder slash menu without needing a dispatch. Two fixes: (1) the cold ACP probe (`acp-probe.ts`) captured `available_commands` but read `probedCommands` synchronously right after `newSession` — racing opencode's async `available_commands_update` notification, so it captured **zero** and only the 7-item static manifest showed. The probe now waits briefly (poll up to 3s for the first batch + a 300ms settle, capped under the 30s probe timeout) so the commands are actually captured. (2) Captured commands are persisted to a new `available_agents.commands` JSONB column and served (merged with the manifest) on the tier-2-probe-skip path, so the agent's discovered commands survive once the model list is warm and show without a dispatch. Boot warms this via the `force: true` startup snapshot. apps/coder only (probe + schema + snapshot). Caveat: depends on opencode emitting `available_commands_update` on session creation rather than only after a prompt — to be confirmed on the host. Claude (PTY) disk/plugin discovery deferred.
+
+## v2.5.9-agent-slash-commands — 2026-05-29
+
+Segmented per-agent slash menu in the coder pane, plus cross-agent skills. The `/` menu now shows two labeled groups — **the active agent's commands first** (opencode/claude/qwen manifest + live ACP `available_commands`), **BooCoder skills second** — instead of always showing BooCoder's skills regardless of provider. `SlashCommandPicker` gains an opt-in `groups` prop (the flat `items` path is unchanged, so **BooChat's menu is byte-identical** — parity verified: no BooChat caller passes the grouped prop, and the skills lookup / invocation routing are untouched); `ChatInput` takes `slashGroups`; `CoderPane` builds the groups from the selected provider's commands + skills. Skills now **run under the selected agent**: the coder `skill_invoke` route accepts a `provider` and, when external, injects the server-side skill body into a dispatched task (instead of native inference) — so a skill like brainstorming executes through opencode/claude with the body kept server-side, mirroring the messages-route external dispatch. Also folds in the earlier initial-chat fix: invoking a skill on the landing chat now runs the same create-chat → assign-to-pane → invoke transition as a text send (`handleLandingSkill`) rather than invoking invisibly without a pane transition (the blank-screen repro). Web tsc + coder build clean.
+
+## v2.5.8-mobile-composer-row — 2026-05-29
+
+Mobile fix for the `AgentComposerBar`: the refresh button was wrapping to a second line. Root cause was layout order, not width — the status dot carried `ml-auto` (pinned to the far-right edge) and the refresh button followed it in DOM order, so it overflowed and wrapped. The dot + refresh are now one right-aligned (`ml-auto`) unit, keeping the refresh on the top line. Additionally, `CompactPicker` gained an `iconOnly` option and the Mode (permission) picker now renders icon-only on mobile (shield + chevron, no "Bypass"/"Plan" text label; `aria-label`/`title` and the tap-to-open list still convey the value) to free row width. Desktop is unchanged (full labels). Web-only change.
+
+## v2.5.7-claude-models-and-picker-fix — 2026-05-29
+
+Two provider-layer changes. **(1) Fix the empty provider picker** — a regression from `v2.5.5` (Phase 2): on a cache miss `getProviderSnapshot` returned synchronous `installed:false` `loading` entries, which `AgentComposerBar` filters out (`e.installed && e.status !== 'error'`); with the client-side poll deferred to Phase 5, a single fetch landed on `loading` forever and no providers appeared. `getProviderSnapshot` now awaits the build and returns terminal entries (the sync `loading` return is deferred until Phase 5 ships the poll); builds stay fast via the tier-2 cold-probe skip. **(2) Claude models** — the list was a hardcoded 2-entry static list (Opus 4 / Sonnet 4, May 2025), and the v2.3 config schema's `models`/`additionalModels` were parsed but never wired. `buildResolvedRegistry` now carries config `models` (replace) + `additionalModels` (merge) onto `ResolvedProviderDef`, and `provider-snapshot` applies them to every ready model list — so `/data/coder-providers.json` can add or replace any provider's models with no code change. Claude `staticModels` bumped to `opus`/`sonnet`/`haiku` latest-aliases plus pinned `claude-opus-4-8` / `claude-sonnet-4-6` / `claude-haiku-4-5-20251001` (passed verbatim to `claude --model`; the CLI accepts both aliases and pinned full names). +2 unit tests (109 total). Builds on `v2.5.6-provider-lifecycle-phase3`.
+
 ## v2.5.6-provider-lifecycle-phase3 — 2026-05-29
 
 Phase 3 of the v2.3 provider-lifecycle batch (`openspec/changes/v2-3-provider-lifecycle/design.md` §5): generic ACP dispatch. `acp-spawn.ts` gains `resolveLaunchSpec(resolved, installPath)` — it consults the resolved registry's `launchCommand` (a config override or a custom-ACP entry's command) first, falling back to the kept `resolveAcpSpawnArgs` switch for built-ins. `acp-dispatch.ts` now spawns `spec.binary`/`spec.args` with `env: { ...process.env, ...spec.env }` instead of the hardcoded per-name argv, and `dispatcher.ts` loads the resolved def by `task.agent` and passes it through. This lets config-defined custom ACP providers dispatch with no new switch case. Built-in dispatch (claude/opencode/goose/qwen) is **byte-identical** to pre-v2.3 — proven by a regression test asserting opencode→`['acp']`, goose→`['acp']`, qwen→`['--acp']`, binary=`installPath ?? id`, and empty config env → plain `process.env`. One deliberate deviation from the spec's literal `!installPath → null`: the `installPath ?? id` fallback is preserved so a missing install path still spawns the bare agent name as before. `setSessionMode`/permission/streaming and the dispatcher poll/NOTIFY/running-guard are untouched. 7 new `acp-spawn.test.ts` cases. No routes/UI (Phase 4+). Builds on `v2.5.5-provider-lifecycle-phase2`.

CLAUDE.md

@@ -68,9 +68,9 @@ Key services:
 - **`messages_with_parts` view** (v1.13.1-B; `schema.sql`). Read sites that need `tool_calls` / `tool_results` / `reasoning_parts` SELECT from this view, NOT `messages` directly. v1.13.20 dropped the legacy `messages.tool_calls` / `messages.tool_results` JSON columns; the view now reads parts-only subselects. Writes target `message_parts` exclusively via `insertParts` (or via the helpers `partsFromAssistantMessage` / `partsFromToolMessage`). The `Message` wire type still carries `tool_calls?` / `tool_results?` because the view synthesizes them from parts — frontend reads are unchanged. Shapes: `tool_calls jsonb[]`, `tool_results jsonb` single object, `reasoning_parts jsonb[]` of `{text}`. If you ever need to UPDATE a message and return its full Message shape, do a two-step UPDATE returning `id` followed by SELECT from the view — RETURNING off the bare `messages` table no longer carries the tool fields.
 - **`services/file_ops.ts`** — Shared file operation implementations used by both inference tools and HTTP routes.
 - **`services/auto_name.ts`** — Non-streaming LLM call to generate 4-word session titles after first assistant reply.
-- **`services/provider-registry.ts`** — Static registry of provider metadata (label, transport, model source). `PROVIDERS` array, `PROVIDERS_BY_NAME` map. 5 providers: boocode (native), opencode (acp), goose (pty), claude (pty), qwen (pty).
+- **`apps/coder/src/services/provider-registry.ts`** (BooCoder, NOT apps/server) — Static registry of provider metadata (label, transport, model source). `PROVIDERS` array, `PROVIDERS_BY_NAME` map. 5 providers: boocode (native), opencode (acp), goose (pty), claude (pty), qwen (pty).
-- **`services/agent-probe.ts`** — Startup probe using direct `exec()` (not SSH). Discovers installed agents on host, their versions, ACP support, and models. Qwen models read from `~/.qwen/settings.json`. Claude models are static from the registry. Results persisted to `available_agents` table.
+- **`apps/coder/src/services/agent-probe.ts`** (BooCoder) — Startup probe using direct `exec()` (not SSH). Discovers installed agents on host, their versions, ACP support, and models. Qwen models read from `~/.qwen/settings.json`. Claude models are static from the registry. Results persisted to `available_agents` table.
-- **`routes/providers.ts`** — `GET /api/providers` returns installed providers with models. Transport field reflects actual capability (checks `supports_acp` from DB, not just registry preference).
+- **`apps/coder/src/routes/providers.ts`** (BooCoder) — `GET /api/providers` returns installed providers with models. Transport field reflects actual capability (checks `supports_acp` from DB, not just registry preference). The apps/server side of this flow is the "Provider picker dispatch" bullet below.
 - **Provider picker dispatch**: when `provider !== 'boocode'`, the message route creates a `tasks` row (with `session_id` set) instead of calling `inference.enqueue`. The dispatcher picks it up and dispatches via ACP or PTY using the agent's `install_path`.
 
 Route registration: all routes registered in `index.ts` via `register*Routes(app, sql, ...)` functions. Routes are in `routes/*.ts`.
@@ -80,12 +80,19 @@ Route registration: all routes registered in `index.ts` via `register*Routes(app
 - Write-capable coding agent. Runs as a **systemd service on the host** (`boocoder.service`), NOT in Docker. Fastify server at port 9502, connects to postgres at `127.0.0.1:5500`.
 - **Workspace dependency on `@boocode/server`**: imports `createInferenceRunner`, `createBroker`, `ALL_TOOLS`, `appendMcpTools` from the server's compiled `dist/`. apps/server's `package.json` has an `exports` map with `types` conditions for NodeNext resolution. apps/server must build FIRST.
 - Build + deploy: `pnpm -C apps/server build && pnpm -C apps/coder build && sudo systemctl restart boocoder`. Env file at `apps/coder/.env.host`. Service file at `/etc/systemd/system/boocoder.service`.
+- After `pnpm -C apps/coder build` the host `boocoder.service` keeps running the OLD process until `sudo systemctl restart boocoder` — a stale process shows **new routes 404 with `{error:'not found'}` while old routes still 200** (the `/api` not-found handler returns that shape). Restart, don't re-debug.
 - Agent dispatch spawns binaries directly using `install_path` from `available_agents` — no `spawn('sh', ['-c', ...])` (fails under systemd). Follows Paseo's pattern: `spawn(fullBinaryPath, argsArray, { cwd })`.
 - systemd hardening: only `NoNewPrivileges=true` is safe. `ProtectSystem`, `ProtectHome`, `PrivateTmp` all break agent dispatch (agents need full filesystem access to read configs, write to worktrees).
 - `apps/server/tsconfig.json` has `declaration: true` so `.d.ts` files exist for workspace consumers.
 - Write tools (`edit_file`, `create_file`, `delete_file`, `apply_pending`, `rewind`) queue in `pending_changes` table. Nothing hits disk until `apply_pending` is called. `write_guard.ts` validates paths (resolve + prefix-check, no realpath since files may not exist for creates).
 - Frontend: NOT a separate SPA. BooCoder is a `'coder'` pane type within BooChat's SPA (`apps/web/`). `CoderPane.tsx` in `apps/web/src/components/panes/`. API requests go through `/api/coder/*` proxy (Vite dev + Fastify production) which rewrites to the boocoder host service (`BOOCODER_URL` env var, default `http://100.114.205.53:9502`). WS connects directly to `:9502`.
 - `apps/coder/web/` is a STANDALONE fallback SPA served at `:9502` directly. The PRIMARY BooCoder frontend is the `CoderPane` in BooChat's SPA (`apps/web/src/components/panes/CoderPane.tsx`), accessible via the "Coder" pane in the workspace at `code.indifferentketchup.com`. Both exist; the pane is what Sam uses.
+- **Provider snapshot lifecycle** (`apps/coder/src/services/`): `provider-config.ts` (Zod config, never-throws on bad input) → `provider-config-registry.ts` (`buildResolvedRegistry`, singleton) → `provider-snapshot.ts` (two-tier probe: tier-1 fast presence, tier-2 cold ACP probe skipped unless force / stale `PROVIDER_PROBE_TTL_MS` 24h / dbEmpty; cached). Verify live: `curl http://100.114.205.53:9502/api/providers/snapshot` — returns providers + models + commands, the exact shape `AgentComposerBar` renders.
+- `PATCH /api/providers/config` replaces a provider id's override object **wholesale** (per-id shallow merge) — to flip one field send `{...existing, enabled}`, or a custom ACP entry's `command`/`label` is wiped and it drops out of the resolved registry. `data/coder-providers.json` is **gitignored** (it's live runtime config — the coder reads AND writes it on UI toggles); the tracked reference is `data/coder-providers.example.json`. The loader falls back to `{providers:{}}` (built-ins only) when the live file is absent, so a fresh checkout needs no copy.
+- **opencode** runs as a warm HTTP server (v2.6 Phase 1, `services/backends/opencode-server.ts` — `opencode serve` per BooCoder process, one opencode session per BooCode session, resumed via `agent_sessions`). goose/qwen/claude still dispatch **one-shot** ACP/PTY with no ctx/token usage; only native `boocode` (llama-swap engine) tracks ctx. Paseo's per-provider native clients (design §12) deliberately not ported.
+- **opencode SSE** (`opencode-server.ts`): live streaming arrives as `session.next.text.delta` / `session.next.reasoning.delta` / `session.next.tool.{called,success,failed}` — NOT `message.part.*` (those are terminal/post-hoc). `client.event.subscribe({ directory })` MUST pass the session's worktree directory; omit it and opencode scopes events to the server's `process.cwd()` → zero session events (empty turns, 180s watchdog timeout). One SSE stream at a time scoped to the last session's dir — concurrent opencode sessions in different worktrees collide (known Phase 1 limit, warns). Turn completes on `session.idle`; `promptAsync` is fire-and-forget (204).
+- **opencode model strings** must be provider-prefixed (`llama-swap/<model>`) AND exist in `~/.config/opencode/opencode.json` `provider.llama-swap.models` — not merely loadable by llama-swap. `parseModel` infers `llama-swap/` for a bare id; the dispatcher coalesces empty→DEFAULT_MODEL then prefixes. `agent-probe` populates opencode's `available_agents.models` via `mergeLlamaSwap` (fetches `/v1/models`); empty model list → frontend sends `''` → no inference (`input:0`, empty turn).
+- **agent_sessions resume**: `config_hash = sha256('opencode_server|<model>')` — must NOT include the server port (random per boot; including it breaks cross-restart resume). `session_worktrees` + `agent_sessions` FKs to `sessions(id)` are `ON DELETE CASCADE` (else DELETE /api/sessions/:id 500s on FK violation). The `@opencode-ai/sdk` v2 client takes flattened params (`{sessionID, directory, parts, model:{providerID,modelID}}`), imports `createOpencodeClient` from `@opencode-ai/sdk/v2/client`.
 
 ### Frontend (`apps/web/src/`)
 
@@ -145,6 +152,7 @@ BooCoder at port 9502: `curl http://100.114.205.53:9502/api/health`. Runs as `bo
 - Tag naming: `vMAJOR.MINOR.PATCH-slug` (e.g. `v1.13.13-ws-publish`). Monotonic per minor — the slug describes the batch's content so the tag name alone is enough to recall what shipped. No letter suffixes (`-a`/`-b`), no pseudo-ranges (`v1.11.x`), no slug-only sub-versions sharing a number (`v1.13.15-tools` + `-openspec` + `-agentlint` — split into sequential patches instead).
 - `CHANGELOG.md` is the per-tag release log, most-recent on top. When a new tag is created, add a `## <tag> — <YYYY-MM-DD>` section with a 3–6 sentence paragraph summarizing what shipped, drawn from the commit body. Cross-reference other tags by name when the batch builds on, fixes, or pairs with prior work (e.g. "pairs with `v1.13.12-ws-schemas`", "fixed in `v1.13.5-stability-bundle`"). No nested bullets — one paragraph.
 - Deploy: `cd /opt/boocode && docker compose up --build -d` (or `docker compose build --no-cache boocode && docker compose up -d` if you suspect a layer-cache issue).
+- The `boocode` container is `build: .` — it builds web+server from the **working tree**, so uncommitted changes deploy. Web edits are live on the Vite dev server (HMR) but NOT on production (`:9500` / code.indifferentketchup.com) until `docker compose up --build -d boocode`.
 - Git push to Gitea: `GIT_SSH_COMMAND="ssh -i /opt/boocode/secrets/boocode_gitea -o IdentitiesOnly=yes" git push origin <branch>`. The default agent identity is rejected; the in-repo deploy key (`secrets/`, gitignored) is the working one. Transient `Connection reset by peer` retries cleanly after `sleep 5`.
 - Don't accumulate `.bak-*` files. Clean them up in the same batch or immediately after merge.
 - DB-integration tests opt-in via env var: `DATABASE_URL='postgres://boocode:devpass@localhost:5500/boochat' pnpm -C apps/server test`. Host port is 5500 (mapped from `boocode_db:5432`); password is `${POSTGRES_PASSWORD}` from `.env` (`devpass`), NOT the literal in `.env`'s `DATABASE_URL=postgres://boocode:Ketchup1479@boocode_db:5432/...` line. `psql` is not on the host PATH — for an interactive query use `docker exec boocode_db psql -U boocode -d boochat -c "..."`. Pattern: `describe.runIf(!!process.env.DATABASE_URL)(...)` with a `beforeAll` that applies the schema via `sql.unsafe(readFileSync(schemaPath))`. Tests skip cleanly when var is unset. `tool_cost_stats.test.ts` is the reference.
@@ -172,10 +180,12 @@ BooCoder at port 9502: `curl http://100.114.205.53:9502/api/health`. Runs as `bo
 - Discriminated unions for type narrowing: `Pane` (by `kind`), `SessionEvent` (by `type`), `InferenceFrame` (by `type`).
 - **Adding a new WS frame type** requires updating BOTH the server's `InferenceFrame` (loose `type:` union + optional fields in `services/inference/turn.ts`) AND the web `WsFrame` (strict discriminated union in `apps/web/src/api/types.ts`). Server publish is permissive; the frontend type is the wire-format gate. The `'usage'` frame added in v1.12.2 needed both sides; missing the web side silently drops the frame at JSON-parse.
 - shadcn primitives live in `components/ui/`. Don't modify them unless adding a new primitive.
+- `ui/` primitives present: button, card, context-menu, dialog, dropdown-menu, input, label, radio-group, sonner, textarea. No switch/sheet/drawer/badge/checkbox — use a `<button role="switch" aria-checked>` toggle (a hand-rolled `Switch` already lives in `SettingsPane.tsx`) and a Dialog-based panel for "drawers".
 - `inferLanguage()` from `lib/attachments.ts` is the canonical file-extension-to-language map. `CodeBlock.tsx` keeps its own `LANG_MAP` because it also resolves markdown fence names.
 - Two UI event buses: `hooks/sessionEvents.ts` for DB-state events (chat_created, session_updated); `lib/events.ts` for ephemeral UI (`sendToTerminal`, `terminalsRegistry`). Don't merge — different subscriber lifecycles.
 - `vite.config.ts` proxy entries are order-sensitive: more-specific prefixes (`/api/term`, `/ws/term`) must come BEFORE `/api`.
 - Mobile pane URL sync (`Session.tsx`): the `?pane=<id>` effect resets `activePaneIdx` whenever `panes` changes. New-pane creation on mobile must push `?pane=` atomically — `addPaneAndSwitch` is the wrapper that does this. `addSplitPane` returns the new pane id for callers.
+- A scrollable list inside a Dialog on mobile: cap `DialogContent` (`max-h-[85vh]` + `grid-rows-[auto_minmax(0,1fr)_auto]`) and make the list the single scroll region with `overscroll-contain` — otherwise touch-scroll drags the whole fixed modal / chains to the page.
 - xterm.js v5 uses canvas rendering — browser doesn't see xterm's selection; the native right-click menu has no working Copy for terminal text. App keybindings (`Cmd/Ctrl-C`, `Cmd/Ctrl-Shift-C`) are the path.
 - **New tools** live in their own `services/<name>.ts` file (see `web_search.ts`, `web_fetch.ts`) — exports a pure `executeFoo(input, ...deps)` for direct test access plus a `ToolDef` wrapper that `loadConfig()`s its real dependencies. Register the ToolDef in `tools.ts` `ALL_TOOLS` (and `READ_ONLY_TOOL_NAMES` if applicable). Inject `fetcher: typeof fetch = fetch` rather than `vi.spyOn(globalThis, 'fetch')` — cleanup is simpler and the production call site stays unchanged.
 - **Sentinels** are `role='system'` rows with structured `metadata.kind` (`cap_hit`, `doom_loop`). UI-only — `buildMessagesPayload` strips them via `isAnySentinel` so the LLM never sees them. A new kind requires arms in `MessageMetadata` in BOTH `apps/server/src/types/api.ts` AND `apps/web/src/api/types.ts`, plus a render branch in `apps/web/src/components/MessageBubble.tsx`.
@@ -191,6 +201,9 @@ BooCoder at port 9502: `curl http://100.114.205.53:9502/api/health`. Runs as `bo
 - **CoderPane uses ChatInput** (`components/panes/CoderPane.tsx`): shares the same `ChatInput` component as BooChat for full parity — attachments, paste-to-chip, auto-grow textarea, queued messages during send. CoderPane's `sendOneMessage` is the send callback; queued messages drain via `useEffect` when `sending` goes false.
 - **Adding a new `SessionEvent` type**: add the interface, add it to the `SessionEvent` union, add a `case` in `useSidebar.ts` `applyEvent` switch (no-op `return prev` is fine), and subscribe in any hook that needs it (e.g. `useSessionStream` for `refetch_messages`).
 - **BooCoder provider registry** (`apps/coder/src/services/provider-registry.ts`): static list of provider defs (boocode, opencode, goose, claude, qwen). `PROBED_AGENT_NAMES` derives from it. Adding/removing providers means editing this file, not the frontend.
+- **AgentComposerBar filters `e.installed`**: provider snapshot entries with `installed:false` (loading/unavailable) are dropped from the dropdown. `getProviderSnapshot` must await the full build — returning synchronous `loading` placeholders makes every provider vanish (the v2.5.7 "no providers showing up" regression); surfacing loading states needs a client poll.
+- **Coder↔web provider-type parity** (`apps/coder/src/services/provider-types.ts` ↔ `apps/web/src/api/types.ts`): enforced by runtime `provider-types-parity.test.ts` (compile-time cross-import is blocked by TS6307 on web's composite tsconfig). Mirror of the ws-frames parity pattern — edit both copies together or the test fails.
+- **ACP command discovery is async**: `acp-probe.ts` must poll after `newSession` for `available_commands_update` (commands arrive in a later notification; reading synchronously captures 0). PTY providers (claude) instead discover from disk via `claude-command-discovery.ts` (`~/.claude/commands` + `enabledPlugins` `skills/`+`commands/`, bare names, deduped). `AgentCommand.kind` tags `'command'` vs `'skill'`; `CoderPane`'s `slashGroups` splits them into icon'd groups. `SlashCommandPicker`'s `groups?` prop is opt-in — BooChat passes flat `items` (unchanged).
 - **Pane header architecture (mobile vs desktop)**: Desktop coder pane header (BooCode label + [+] [×]) lives in `Workspace.tsx` gated by `isCoder && !isMobile`. Mobile coder controls (● ×) live in `Session.tsx` header row next to `MobileTabSwitcher`/`NewPaneMenu`. `AgentComposerBar` (provider/mode/model pickers) renders inside `CoderPane.tsx` on both. The ● status dot is passed via `connected` prop from CoderPane to AgentComposerBar.
 - **MessageBubble shared between BooChat and BooCoder** (`components/MessageBubble.tsx`): accepts optional `actions?: MessageActions` callbacks (onRegenerate, onResend, onFork, onDelete) and `hideActions?: ('fork'|'delete'|'openInPane')[]`. Defaults use BooChat API; CoderPane overrides via `CoderMessageList` props. `CoderTextBubble` was removed. **`CoderMessageList` passes `CoderMessageWire as unknown as Message`** — the coder wire shape lacks `metadata`/`kind`/`summary`, so those fields are `undefined` (not `null`) on coder messages. Null-guards on any `Message` field MUST use loose `!= null`, not strict `!== null` (`undefined !== null` is `true` → `.kind` throws → blank-screen crash). The `as unknown as` cast hides this from tsc; build + typecheck pass while runtime crashes.
 - **llama-sidecar** (`/opt/forks/llama-sidecar/`): Go daemon for per-agent llama-server process pool. Cross-compile: `GOOS=windows GOARCH=amd64 /snap/go/current/bin/go build -o bin/llama-sidecar.exe ./cmd/llama-sidecar`. Gitea: `indifferentketchup/llama-sidecar`. Windows child process gotchas: use `context.Background()` for child lifetime (not request ctx), `os.Open(os.DevNull)` for stdin, `os.Pipe()` for stdout with drain goroutine, `DETACHED_PROCESS | CREATE_NEW_PROCESS_GROUP` creation flags. SSH to sam-desktop: `ssh samki@100.101.41.16`; use `schtasks` for persistent process spawning (SSH `start /B` doesn't survive session close).

apps/coder/package.json

@@ -16,6 +16,7 @@
     "@agentclientprotocol/sdk": "^0.22.1",
     "@boocode/server": "workspace:*",
     "@fastify/static": "^7.0.4",
+    "@opencode-ai/sdk": "~1.15.0",
     "@fastify/websocket": "^10.0.1",
     "@modelcontextprotocol/sdk": "^1.29.0",
     "fastify": "^4.28.1",

apps/coder/src/index.ts

@@ -33,6 +33,7 @@ import { registerProviderRoutes } from './routes/providers.js';
 import { registerWebSocket } from './routes/ws.js';
 // Phase 4: dispatcher + agent probe
 import { createDispatcher } from './services/dispatcher.js';
+import { agentPool } from './services/agent-pool.js';
 import { probeAgents } from './services/agent-probe.js';
 import { getProviderSnapshot, persistProbedModels } from './services/provider-snapshot.js';
 import { setPermissionHooks } from './services/permission-waiter.js';
@@ -178,7 +179,12 @@ async function main() {
   // Phase 4: dispatcher — polls tasks table and runs inference
   const dispatcher = createDispatcher({ sql, inference: inferenceApi, broker, log: app.log, config });
   dispatcher.start();
-  app.addHook('onClose', () => dispatcher.stop());
+  app.addHook('onClose', async () => {
+    // stop() first so in-flight dispatcher turns settle, then drain the pool.
+    // Pool is empty in Phase 0 (nothing spawns yet) — dispose() is inert.
+    await dispatcher.stop();
+    await agentPool.dispose();
+  });
 
   // Register routes
   registerMessageRoutes(app, sql, broker, inferenceApi);

apps/coder/src/routes/__tests__/providers.routes.test.ts

@@ -0,0 +1,211 @@
+import { describe, it, expect, beforeEach, afterEach, vi } from 'vitest';
+import Fastify, { type FastifyInstance } from 'fastify';
+import { existsSync, readFileSync, writeFileSync, rmSync } from 'node:fs';
+import { tmpdir } from 'node:os';
+import { join } from 'node:path';
+import { registerProviderRoutes } from '../providers.js';
+import { load } from '../../services/provider-config.js';
+import { loadProviderConfig } from '../../services/provider-config-registry.js';
+import { clearProviderSnapshotCache } from '../../services/provider-snapshot.js';
+import type { Config } from '../../config.js';
+import type { Sql } from '../../db.js';
+
+/** Minimal sql stub: available_agents reads return []. */
+function mockSql(): Sql {
+  return vi.fn((strings: TemplateStringsArray) => {
+    const q = strings.join('');
+    if (q.includes('available_agents')) return Promise.resolve([]);
+    return Promise.resolve([]);
+  }) as unknown as Sql;
+}
+
+let tmpCounter = 0;
+function freshPath(): string {
+  tmpCounter += 1;
+  return join(tmpdir(), `coder-providers-routes-${process.pid}-${tmpCounter}.json`);
+}
+
+function buildApp(providersPath: string): FastifyInstance {
+  const app = Fastify();
+  // Mirror index.ts: tolerate empty JSON bodies.
+  app.removeContentTypeParser(['application/json']);
+  app.addContentTypeParser('application/json', { parseAs: 'string' }, (_req, body, done) => {
+    const str = (body as string) ?? '';
+    if (str.trim().length === 0) return done(null, {});
+    try {
+      done(null, JSON.parse(str));
+    } catch (err) {
+      done(err as Error, undefined);
+    }
+  });
+  const config = {
+    CODER_PROVIDERS_PATH: providersPath,
+    LLAMA_SWAP_URL: 'http://llama-swap.test',
+    PROVIDER_PROBE_TTL_MS: 86_400_000,
+  } as unknown as Config;
+  registerProviderRoutes(app, mockSql(), config);
+  return app;
+}
+
+const JSON_HEADERS = { 'content-type': 'application/json' };
+const createdPaths: string[] = [];
+
+beforeEach(() => {
+  clearProviderSnapshotCache();
+  loadProviderConfig('/nonexistent-coder-providers.json'); // reset registry to built-ins
+  vi.restoreAllMocks();
+  vi.stubGlobal('fetch', vi.fn().mockRejectedValue(new Error('no network in test')));
+});
+
+afterEach(() => {
+  for (const p of createdPaths.splice(0)) {
+    try {
+      rmSync(p, { force: true });
+    } catch {
+      /* ignore */
+    }
+  }
+});
+
+describe('GET /api/providers/config', () => {
+  it('returns the current config file (built-ins-only when missing)', async () => {
+    const path = freshPath();
+    createdPaths.push(path);
+    const app = buildApp(path);
+    const res = await app.inject({ method: 'GET', url: '/api/providers/config' });
+    expect(res.statusCode).toBe(200);
+    expect(res.json()).toEqual({ providers: {} });
+    await app.close();
+  });
+
+  it('reflects an existing file', async () => {
+    const path = freshPath();
+    createdPaths.push(path);
+    writeFileSync(path, JSON.stringify({ providers: { goose: { enabled: false } } }));
+    const app = buildApp(path);
+    const res = await app.inject({ method: 'GET', url: '/api/providers/config' });
+    expect(res.json()).toEqual({ providers: { goose: { enabled: false } } });
+    await app.close();
+  });
+});
+
+describe('PATCH /api/providers/config', () => {
+  it('valid patch → 200, writes the merged file (order: validate→save→reload→clear)', async () => {
+    const path = freshPath();
+    createdPaths.push(path);
+    writeFileSync(path, JSON.stringify({ providers: { goose: { label: 'Goose' } } }));
+    const app = buildApp(path);
+
+    const res = await app.inject({
+      method: 'PATCH',
+      url: '/api/providers/config',
+      headers: JSON_HEADERS,
+      payload: JSON.stringify({ providers: { opencode: { enabled: false } } }),
+    });
+
+    expect(res.statusCode).toBe(200);
+    expect(res.json()).toMatchObject({ ok: true });
+    // File written + merged (goose untouched, opencode added).
+    const onDisk = load(path);
+    expect(onDisk.providers).toEqual({
+      goose: { label: 'Goose' },
+      opencode: { enabled: false },
+    });
+    await app.close();
+  });
+
+  it('null value deletes the override', async () => {
+    const path = freshPath();
+    createdPaths.push(path);
+    writeFileSync(path, JSON.stringify({ providers: { goose: { enabled: false }, opencode: { enabled: false } } }));
+    const app = buildApp(path);
+
+    const res = await app.inject({
+      method: 'PATCH',
+      url: '/api/providers/config',
+      headers: JSON_HEADERS,
+      payload: JSON.stringify({ providers: { goose: null } }),
+    });
+
+    expect(res.statusCode).toBe(200);
+    expect(load(path).providers).toEqual({ opencode: { enabled: false } });
+    await app.close();
+  });
+
+  it('INVALID body → 422 and the file is NOT written (validate before save)', async () => {
+    const path = freshPath();
+    createdPaths.push(path);
+    const before = JSON.stringify({ providers: { goose: { enabled: true } } });
+    writeFileSync(path, before);
+    const app = buildApp(path);
+
+    const res = await app.inject({
+      method: 'PATCH',
+      url: '/api/providers/config',
+      headers: JSON_HEADERS,
+      payload: JSON.stringify({ providers: { goose: { enabled: 'yes' } } }), // bad type
+    });
+
+    expect(res.statusCode).toBe(422);
+    // File must be byte-for-byte unchanged — nothing written on a 422.
+    expect(readFileSync(path, 'utf8')).toBe(before);
+    await app.close();
+  });
+
+  it('save failure → 500 and the file is NOT created (no state divergence)', async () => {
+    const path = join(tmpdir(), `no-such-dir-${process.pid}-${Date.now()}`, 'coder-providers.json');
+    const app = buildApp(path);
+
+    const res = await app.inject({
+      method: 'PATCH',
+      url: '/api/providers/config',
+      headers: JSON_HEADERS,
+      payload: JSON.stringify({ providers: { goose: { enabled: false } } }),
+    });
+
+    expect(res.statusCode).toBe(500);
+    expect(existsSync(path)).toBe(false);
+    await app.close();
+  });
+});
+
+describe('POST /api/providers/refresh', () => {
+  it('no body → refreshes all registered providers', async () => {
+    const app = buildApp(freshPath());
+    const res = await app.inject({ method: 'POST', url: '/api/providers/refresh' });
+    expect(res.statusCode).toBe(200);
+    expect(res.json().refreshed).toBeGreaterThan(0);
+    await app.close();
+  });
+
+  it('subset body → refreshed count reflects only the requested providers', async () => {
+    const app = buildApp(freshPath());
+    const res = await app.inject({
+      method: 'POST',
+      url: '/api/providers/refresh',
+      headers: JSON_HEADERS,
+      payload: JSON.stringify({ providers: ['boocode'] }),
+    });
+    expect(res.statusCode).toBe(200);
+    expect(res.json()).toEqual({ refreshed: 1 });
+    await app.close();
+  });
+});
+
+describe('GET /api/providers/:id/diagnostic', () => {
+  it('known provider → 200 JSON { diagnostic }', async () => {
+    const app = buildApp(freshPath());
+    const res = await app.inject({ method: 'GET', url: '/api/providers/boocode/diagnostic' });
+    expect(res.statusCode).toBe(200);
+    expect(res.headers['content-type']).toContain('application/json');
+    expect(res.json().diagnostic).toContain('provider: boocode');
+    await app.close();
+  });
+
+  it('unknown provider → 404', async () => {
+    const app = buildApp(freshPath());
+    const res = await app.inject({ method: 'GET', url: '/api/providers/nope/diagnostic' });
+    expect(res.statusCode).toBe(404);
+    await app.close();
+  });
+});

apps/coder/src/routes/providers.ts

@@ -1,7 +1,29 @@
 import type { FastifyInstance } from 'fastify';
+import { z } from 'zod';
 import type { Sql } from '../db.js';
 import type { Config } from '../config.js';
-import { getProviderSnapshot, clearProviderSnapshotCache } from '../services/provider-snapshot.js';
+import {
+  getProviderSnapshot,
+  clearProviderSnapshotCache,
+  peekSnapshotEntry,
+} from '../services/provider-snapshot.js';
+import {
+  load,
+  save,
+  CoderProvidersFileSchema,
+  ProviderConfigPatchSchema,
+  mergeProviderConfigPatch,
+} from '../services/provider-config.js';
+import {
+  reloadProviderConfig,
+  getResolvedRegistry,
+} from '../services/provider-config-registry.js';
+import {
+  getProviderDiagnostic,
+  type DiagnosticAgentRow,
+} from '../services/provider-diagnostic.js';
+
+const RefreshBodySchema = z.object({ providers: z.array(z.string()).optional() });
 
 export function registerProviderRoutes(app: FastifyInstance, sql: Sql, config: Config): void {
   app.get<{ Querystring: { cwd?: string } }>('/api/providers/snapshot', async (req, _reply) => {
@@ -9,9 +31,97 @@ export function registerProviderRoutes(app: FastifyInstance, sql: Sql, config: C
     return getProviderSnapshot(sql, config, cwd);
   });
 
-  app.post('/api/providers/refresh', async (_req, _reply) => {
+  // 4.1 — current loaded config file (raw CoderProvidersFile, not the resolved registry).
+  app.get('/api/providers/config', async (_req, _reply) => {
+    return load(config.CODER_PROVIDERS_PATH);
+  });
+
+  // 4.2 — patch the config file (design.md §6.2). Strict order is the whole
+  // correctness story: validate → save → reload → clear. A malformed body or an
+  // invalid merged result returns 422 and NEVER writes; a save failure returns
+  // 500 and leaves in-memory state untouched (no file/registry divergence).
+  app.patch('/api/providers/config', async (req, reply) => {
+    // 1. Validate the PATCH body shape (malformed → 422, never reaches merge).
+    const parsed = ProviderConfigPatchSchema.safeParse(req.body);
+    if (!parsed.success) {
+      return reply.code(422).send({
+        error: 'invalid provider config patch',
+        issues: parsed.error.flatten(),
+      });
+    }
+
+    // 2. Shallow per-id merge over the current file (null deletes; object replaces).
+    const current = load(config.CODER_PROVIDERS_PATH);
+    const merged = mergeProviderConfigPatch(current, parsed.data);
+
+    // 3. Validate the merged result — refuse to write a config that won't load.
+    const validated = CoderProvidersFileSchema.safeParse(merged);
+    if (!validated.success) {
+      return reply.code(422).send({
+        error: 'merged provider config is invalid',
+        issues: validated.error.flatten(),
+      });
+    }
+
+    // 4. Persist. If save throws, STOP here — do NOT reload/clear, so the file on
+    // disk and the in-memory resolved registry can never diverge.
+    try {
+      save(config.CODER_PROVIDERS_PATH, validated.data);
+    } catch (err) {
+      req.log.error(
+        { err: err instanceof Error ? err.message : String(err), path: config.CODER_PROVIDERS_PATH },
+        'provider-config: save failed — in-memory state untouched',
+      );
+      return reply.code(500).send({ error: 'failed to write provider config' });
+    }
+
+    // 5 + 6. Rebuild the in-memory resolved registry from the new file, then drop
+    // the snapshot cache so the next /snapshot reflects the change.
+    reloadProviderConfig();
+    clearProviderSnapshotCache();
+
+    // 7. Return the new config (per §6.2 `{ ok: true }`, plus the merged providers
+    // so the client can update without a follow-up GET).
+    return { ok: true, providers: validated.data.providers };
+  });
+
+  // 4.3 — force a cold probe. Optional { providers?: string[] } narrows the
+  // reported subset (design.md §6.3 Paseo pattern). The force=true snapshot is
+  // the only existing re-probe primitive (per-provider force would be a
+  // snapshot-internal change, out of Phase 4 scope), so the probe runs for all
+  // installed providers; the `refreshed` count reflects the requested subset.
+  app.post('/api/providers/refresh', async (req, reply) => {
+    const parsed = RefreshBodySchema.safeParse(req.body ?? {});
+    if (!parsed.success) {
+      return reply.code(422).send({ error: 'invalid refresh body', issues: parsed.error.flatten() });
+    }
+    const subset = parsed.data.providers;
     clearProviderSnapshotCache();
     const entries = await getProviderSnapshot(sql, config, undefined, true);
-    return { refreshed: entries.length };
+    const refreshed =
+      subset && subset.length > 0
+        ? entries.filter((e) => subset.includes(e.name)).length
+        : entries.length;
+    return { refreshed };
+  });
+
+  // 4.4 — per-provider diagnostic (design.md §6.4 → JSON `{ diagnostic: string }`).
+  // Read-only: reports cached state (resolved def + available_agents row + warm
+  // snapshot cache for the last probe error) plus a `which` PATH check. No probe
+  // spawn. The report itself is a plaintext block (§8); the route wraps it as JSON.
+  app.get<{ Params: { id: string } }>('/api/providers/:id/diagnostic', async (req, reply) => {
+    const id = req.params.id;
+    const resolved = getResolvedRegistry().get(id);
+    if (!resolved) {
+      return reply.code(404).send({ error: `unknown provider '${id}'` });
+    }
+    const rows = await sql<DiagnosticAgentRow[]>`
+      SELECT name, install_path, supports_acp, models, last_probed_at
+      FROM available_agents WHERE name = ${id}
+    `;
+    const report = await getProviderDiagnostic(resolved, rows[0], {
+      cachedEntry: peekSnapshotEntry(id),
+    });
+    return { diagnostic: report };
   });
 }

apps/coder/src/routes/skills.ts

@@ -16,6 +16,12 @@ const SkillInvokeBody = z.object({
   pane_id: z.string().min(1).max(200),
   skill_name: z.string().min(1),
   user_message: z.string().max(64_000).nullable().optional(),
+  // v2.5.9: when set to an external provider, the skill runs UNDER that agent —
+  // its body is injected into a dispatched task instead of native inference.
+  provider: z.string().max(100).optional(),
+  model: z.string().max(200).optional(),
+  mode_id: z.string().max(200).optional(),
+  thinking_option_id: z.string().max(200).optional(),
 });
 
 interface InferenceApi {
@@ -39,9 +45,9 @@ export function registerSkillRoutes(
       }
 
       const sessionId = req.params.sessionId;
-      const { pane_id, skill_name } = parsed.data;
+      const { pane_id, skill_name, provider, model, mode_id, thinking_option_id } = parsed.data;
-      const sessionRows = await sql<{ id: string }[]>`
+      const sessionRows = await sql<{ id: string; project_id: string }[]>`
-        SELECT id FROM sessions WHERE id = ${sessionId}
+        SELECT id, project_id FROM sessions WHERE id = ${sessionId}
       `;
       if (sessionRows.length === 0) {
         reply.code(404);
@@ -69,6 +75,31 @@ export function registerSkillRoutes(
         return { error: 'unknown_skill', message: `unknown skill: ${skill_name}` };
       }
 
+      // v2.5.9: external agent → run the skill UNDER that agent. The skill body
+      // stays server-side (like the native path's tool message) and is injected
+      // into a dispatched task; the agent receives the skill instructions + the
+      // user's text. Mirrors the messages-route external-provider dispatch.
+      if (provider && provider !== 'boocode') {
+        const [userMsg] = await sql<{ id: string }[]>`
+          INSERT INTO messages (session_id, chat_id, role, content, status, created_at)
+          VALUES (${sessionId}, ${chatId}, 'user', ${userText}, 'complete', clock_timestamp())
+          RETURNING id
+        `;
+        broker.publishFrame(sessionId, { type: 'message_started', message_id: userMsg!.id, chat_id: chatId, role: 'user' } as WsFrame);
+        broker.publishFrame(sessionId, { type: 'delta', message_id: userMsg!.id, chat_id: chatId, content: userText } as WsFrame);
+        broker.publishFrame(sessionId, { type: 'message_complete', message_id: userMsg!.id, chat_id: chatId } as WsFrame);
+
+        const taskInput = `${body}\n\n---\n\n${userText}`;
+        const [task] = await sql<{ id: string; state: string }[]>`
+          INSERT INTO tasks (project_id, input, agent, model, mode_id, thinking_option_id, session_id)
+          VALUES (${sessionRows[0]!.project_id}, ${taskInput}, ${provider}, ${model ?? null}, ${mode_id ?? null}, ${thinking_option_id ?? null}, ${sessionId})
+          RETURNING id, state
+        `;
+        await sql`UPDATE chats SET updated_at = clock_timestamp() WHERE id = ${chatId}`;
+        reply.code(202);
+        return { user_message_id: userMsg!.id, task_id: task!.id, dispatched: true };
+      }
+
       const { result, toolCall } = await runSkillInvokeTransaction(sql, {
         sessionId,
         chatId,

apps/coder/src/schema.sql

@@ -66,12 +66,70 @@ CREATE OR REPLACE VIEW human_inbox AS
 ALTER TABLE available_agents ADD COLUMN IF NOT EXISTS models JSONB DEFAULT '[]'::jsonb;
 ALTER TABLE available_agents ADD COLUMN IF NOT EXISTS label TEXT;
 ALTER TABLE available_agents ADD COLUMN IF NOT EXISTS transport TEXT DEFAULT 'pty';
+-- v2.5.10: persisted ACP available_commands (captured during the cold probe), so
+-- an agent's live command set survives the tier-2 probe skip and shows without a
+-- dispatch.
+ALTER TABLE available_agents ADD COLUMN IF NOT EXISTS commands JSONB DEFAULT '[]'::jsonb;
 
 -- v2.2.0: Paseo-style session config on tasks.
 ALTER TABLE tasks ADD COLUMN IF NOT EXISTS mode_id TEXT;
 ALTER TABLE tasks ADD COLUMN IF NOT EXISTS thinking_option_id TEXT;
 ALTER TABLE tasks ADD COLUMN IF NOT EXISTS feature_values JSONB;
 
+-- v2.6: one shared worktree per session (all agents/panes in the session operate in it).
+CREATE TABLE IF NOT EXISTS session_worktrees (
+  session_id UUID PRIMARY KEY REFERENCES sessions(id) ON DELETE CASCADE,
+  worktree_path TEXT NOT NULL,
+  base_commit TEXT,
+  created_at TIMESTAMPTZ NOT NULL DEFAULT clock_timestamp()
+);
+-- Migrate existing FK to CASCADE (idempotent: drops the old constraint if present).
+DO $$ BEGIN
+  IF EXISTS (
+    SELECT 1 FROM pg_constraint
+    WHERE conname = 'session_worktrees_session_id_fkey'
+      AND confdeltype <> 'c'
+  ) THEN
+    ALTER TABLE session_worktrees DROP CONSTRAINT session_worktrees_session_id_fkey;
+    ALTER TABLE session_worktrees ADD CONSTRAINT session_worktrees_session_id_fkey
+      FOREIGN KEY (session_id) REFERENCES sessions(id) ON DELETE CASCADE;
+  END IF;
+END $$;
+
+-- v2.6: one backend session per (session, agent); resumed on switch-back.
+CREATE TABLE IF NOT EXISTS agent_sessions (
+  session_id UUID NOT NULL REFERENCES sessions(id) ON DELETE CASCADE,
+  agent TEXT NOT NULL,
+  backend TEXT NOT NULL,
+  agent_session_id TEXT,
+  server_port INTEGER,
+  status TEXT NOT NULL DEFAULT 'idle',
+  last_active_at TIMESTAMPTZ,
+  created_at TIMESTAMPTZ NOT NULL DEFAULT clock_timestamp(),
+  PRIMARY KEY (session_id, agent),
+  CONSTRAINT agent_sessions_backend_chk CHECK (backend IN ('opencode_server', 'acp_warm')),
+  CONSTRAINT agent_sessions_status_chk CHECK (status IN ('idle', 'active', 'crashed', 'closed'))
+);
+
+-- Migrate existing agent_sessions FK to CASCADE.
+DO $$ BEGIN
+  IF EXISTS (
+    SELECT 1 FROM pg_constraint
+    WHERE conname = 'agent_sessions_session_id_fkey'
+      AND confdeltype <> 'c'
+  ) THEN
+    ALTER TABLE agent_sessions DROP CONSTRAINT agent_sessions_session_id_fkey;
+    ALTER TABLE agent_sessions ADD CONSTRAINT agent_sessions_session_id_fkey
+      FOREIGN KEY (session_id) REFERENCES sessions(id) ON DELETE CASCADE;
+  END IF;
+END $$;
+
+-- v2.6: config fingerprint for stale-session detection (auto-recover on model change).
+ALTER TABLE agent_sessions ADD COLUMN IF NOT EXISTS config_hash TEXT;
+
+-- v2.6: attribution for DiffPanel badges (Phase 1 UX reads this).
+ALTER TABLE pending_changes ADD COLUMN IF NOT EXISTS agent TEXT;
+
 -- LISTEN/NOTIFY fast path: every tasks INSERT (from any call site — routes,
 -- new_task tool, arena, MCP server) fires pg_notify('tasks_new') in the same
 -- transaction, so the dispatcher reacts immediately instead of waiting for the

apps/coder/src/services/__tests__/acp-client-fs.test.ts

@@ -0,0 +1,50 @@
+import { describe, it, expect, afterEach } from 'vitest';
+import { mkdtempSync, rmSync } from 'node:fs';
+import { tmpdir } from 'node:os';
+import { join } from 'node:path';
+import { readWorktreeTextFile, writeWorktreeTextFile } from '../acp-client-fs.js';
+
+const created: string[] = [];
+function freshWorktree(): string {
+  const wt = mkdtempSync(join(tmpdir(), 'acp-wt-'));
+  created.push(wt);
+  return wt;
+}
+afterEach(() => {
+  for (const d of created.splice(0)) {
+    try {
+      rmSync(d, { recursive: true, force: true });
+      rmSync(`${d}-evil`, { recursive: true, force: true });
+    } catch {
+      /* ignore */
+    }
+  }
+});
+
+describe('acp-client-fs worktree scoping', () => {
+  it('writes then reads a file inside the worktree', async () => {
+    const wt = freshWorktree();
+    await writeWorktreeTextFile(wt, 'sub/dir/note.txt', 'hello');
+    expect(await readWorktreeTextFile(wt, 'sub/dir/note.txt')).toBe('hello');
+  });
+
+  it('rejects ../ traversal on read', async () => {
+    const wt = freshWorktree();
+    await expect(readWorktreeTextFile(wt, '../../etc/passwd')).rejects.toThrow(/escapes worktree/);
+  });
+
+  it('rejects ../ traversal on write', async () => {
+    const wt = freshWorktree();
+    await expect(writeWorktreeTextFile(wt, '../escape.txt', 'x')).rejects.toThrow(/escapes worktree/);
+  });
+
+  it('rejects a sibling-prefix path (the unbounded-startsWith bug)', async () => {
+    const wt = freshWorktree();
+    // Absolute path that shares the worktree as a STRING prefix but is a sibling
+    // dir: `<wt>-evil/...`. A bare `startsWith(<wt>)` wrongly admits it.
+    await expect(readWorktreeTextFile(wt, `${wt}-evil/secret.txt`)).rejects.toThrow(/escapes worktree/);
+    await expect(writeWorktreeTextFile(wt, `${wt}-evil/secret.txt`, 'x')).rejects.toThrow(
+      /escapes worktree/,
+    );
+  });
+});

apps/coder/src/services/__tests__/provider-config-registry.test.ts

@@ -61,6 +61,22 @@ describe('buildResolvedRegistry', () => {
     warn.mockRestore();
   });
 
+  it('carries config models + additionalModels onto built-in and custom defs', () => {
+    const reg = buildResolvedRegistry(PROVIDERS, {
+      providers: {
+        claude: { models: [{ id: 'claude-opus-4-8', label: 'Opus 4.8' }] },
+        'amp-acp': {
+          extends: 'acp',
+          label: 'Amp',
+          command: ['amp-acp'],
+          additionalModels: [{ id: 'amp-1', label: 'Amp 1' }],
+        },
+      },
+    });
+    expect(reg.get('claude')!.configModels).toEqual([{ id: 'claude-opus-4-8', label: 'Opus 4.8' }]);
+    expect(reg.get('amp-acp')!.configAdditionalModels).toEqual([{ id: 'amp-1', label: 'Amp 1' }]);
+  });
+
   it('REGRESSION: empty config returns exactly the built-ins, all enabled', () => {
     const reg = buildResolvedRegistry(PROVIDERS, { providers: {} });
     expect(reg.size).toBe(PROVIDERS.length);

apps/coder/src/services/__tests__/provider-config.test.ts

@@ -0,0 +1,96 @@
+import { describe, it, expect } from 'vitest';
+import {
+  mergeProviderConfigPatch,
+  ProviderConfigPatchSchema,
+  CoderProvidersFileSchema,
+  type CoderProvidersFile,
+} from '../provider-config.js';
+
+describe('ProviderConfigPatchSchema', () => {
+  it('accepts a per-provider override patch', () => {
+    const parsed = ProviderConfigPatchSchema.safeParse({ providers: { goose: { enabled: false } } });
+    expect(parsed.success).toBe(true);
+  });
+
+  it('accepts a null value (delete-the-override sentinel)', () => {
+    const parsed = ProviderConfigPatchSchema.safeParse({ providers: { goose: null } });
+    expect(parsed.success).toBe(true);
+  });
+
+  it('defaults providers to {} on an empty body', () => {
+    const parsed = ProviderConfigPatchSchema.safeParse({});
+    expect(parsed.success).toBe(true);
+    if (parsed.success) expect(parsed.data.providers).toEqual({});
+  });
+
+  it('rejects a malformed override (wrong field type)', () => {
+    const parsed = ProviderConfigPatchSchema.safeParse({ providers: { goose: { enabled: 'yes' } } });
+    expect(parsed.success).toBe(false);
+  });
+
+  it('rejects a non-object providers map', () => {
+    const parsed = ProviderConfigPatchSchema.safeParse({ providers: 123 });
+    expect(parsed.success).toBe(false);
+  });
+});
+
+describe('mergeProviderConfigPatch', () => {
+  const current: CoderProvidersFile = {
+    providers: {
+      goose: { enabled: true, label: 'Goose' },
+      opencode: { enabled: true },
+    },
+  };
+
+  it('replaces an existing override object wholesale (not deep-merge)', () => {
+    const merged = mergeProviderConfigPatch(current, { providers: { goose: { enabled: false } } });
+    // Whole override replaced — the prior `label` is gone, only `enabled` remains.
+    expect(merged.providers.goose).toEqual({ enabled: false });
+  });
+
+  it('adds a brand-new override id', () => {
+    const merged = mergeProviderConfigPatch(current, {
+      providers: { 'amp-acp': { extends: 'acp', label: 'Amp', command: ['amp-acp'] } },
+    });
+    expect(merged.providers['amp-acp']).toEqual({ extends: 'acp', label: 'Amp', command: ['amp-acp'] });
+  });
+
+  it('deletes an override when the value is null', () => {
+    const merged = mergeProviderConfigPatch(current, { providers: { goose: null } });
+    expect(merged.providers.goose).toBeUndefined();
+    expect(Object.keys(merged.providers)).toEqual(['opencode']);
+  });
+
+  it('leaves ids absent from the patch untouched', () => {
+    const merged = mergeProviderConfigPatch(current, { providers: { goose: { enabled: false } } });
+    expect(merged.providers.opencode).toEqual({ enabled: true });
+  });
+
+  it('does not mutate the input config', () => {
+    const snapshot = JSON.parse(JSON.stringify(current));
+    mergeProviderConfigPatch(current, { providers: { goose: null, opencode: { enabled: false } } });
+    expect(current).toEqual(snapshot);
+  });
+
+  it('empty patch returns an equivalent config', () => {
+    const merged = mergeProviderConfigPatch(current, { providers: {} });
+    expect(merged).toEqual(current);
+  });
+});
+
+describe('CoderProvidersFileSchema (validate-before-save guard)', () => {
+  it('accepts a clean merged config', () => {
+    const merged = mergeProviderConfigPatch(
+      { providers: {} },
+      { providers: { goose: { enabled: false } } },
+    );
+    expect(CoderProvidersFileSchema.safeParse(merged).success).toBe(true);
+  });
+
+  it('rejects a config carrying an invalid override (never written)', () => {
+    // A merged object that somehow holds a bad override must fail validation
+    // so the PATCH route returns 422 and never calls save().
+    const invalid = { providers: { goose: { enabled: 'nope' } } };
+    expect(CoderProvidersFileSchema.safeParse(invalid).success).toBe(false);
+  });
+});

apps/coder/src/services/__tests__/provider-diagnostic.test.ts

@@ -0,0 +1,85 @@
+import { describe, it, expect } from 'vitest';
+import { getProviderDiagnostic, type DiagnosticAgentRow } from '../provider-diagnostic.js';
+import { buildResolvedRegistry } from '../provider-config-registry.js';
+import { PROVIDERS } from '../provider-registry.js';
+import type { ProviderSnapshotEntry } from '../provider-types.js';
+
+const registry = buildResolvedRegistry(PROVIDERS, {
+  providers: {
+    goose: { enabled: false },
+    'amp-acp': { extends: 'acp', label: 'Amp', command: ['amp-acp', '--acp'] },
+  },
+});
+
+const alwaysAvailable = () => Promise.resolve(true);
+const neverAvailable = () => Promise.resolve(false);
+
+describe('getProviderDiagnostic', () => {
+  it('reports a disabled built-in (enabled:false, no install)', async () => {
+    const report = await getProviderDiagnostic(registry.get('goose')!, undefined, {
+      checkAvailable: neverAvailable,
+    });
+    expect(report).toContain('provider: goose');
+    expect(report).toContain('enabled: false');
+    expect(report).toContain('installed: false');
+    expect(report).toMatch(/command_available:\s*false/);
+  });
+
+  it('reports an installed built-in with its install_path, last_probed_at, model count', async () => {
+    const agentRow: DiagnosticAgentRow = {
+      name: 'opencode',
+      install_path: '/usr/bin/opencode',
+      supports_acp: true,
+      models: [
+        { id: 'm1', label: 'M1' },
+        { id: 'm2', label: 'M2' },
+      ],
+      last_probed_at: '2026-05-29T12:00:00.000Z',
+    };
+    const report = await getProviderDiagnostic(registry.get('opencode')!, agentRow, {
+      checkAvailable: alwaysAvailable,
+    });
+    expect(report).toContain('install_path: /usr/bin/opencode');
+    expect(report).toContain('2026-05-29T12:00:00.000Z');
+    expect(report).toContain('installed: true');
+    expect(report).toMatch(/models_in_db:\s*2/);
+    expect(report).toMatch(/command_available:\s*true/);
+  });
+
+  it('reports a custom ACP launch command + its binary', async () => {
+    const report = await getProviderDiagnostic(registry.get('amp-acp')!, undefined, {
+      checkAvailable: alwaysAvailable,
+    });
+    expect(report).toContain('provider: amp-acp');
+    expect(report).toContain('amp-acp --acp');
+    expect(report).toContain('customAcp: true');
+  });
+
+  it('surfaces the last probe error from a cached snapshot entry', async () => {
+    const cachedEntry: ProviderSnapshotEntry = {
+      name: 'opencode',
+      label: 'OpenCode',
+      transport: 'acp',
+      status: 'error',
+      enabled: true,
+      installed: true,
+      models: [],
+      modes: [],
+      defaultModeId: null,
+      commands: [],
+      error: 'ACP initialize timed out',
+    };
+    const report = await getProviderDiagnostic(registry.get('opencode')!, undefined, {
+      cachedEntry,
+      checkAvailable: alwaysAvailable,
+    });
+    expect(report).toContain('ACP initialize timed out');
+  });
+
+  it('reports no error when none is cached', async () => {
+    const report = await getProviderDiagnostic(registry.get('opencode')!, undefined, {
+      checkAvailable: alwaysAvailable,
+    });
+    expect(report).toMatch(/last_probe_error:\s*\(none/);
+  });
+});

apps/coder/src/services/__tests__/provider-snapshot.test.ts

@@ -7,6 +7,7 @@ import {
   prefixLlamaSwapModels,
   clearProviderSnapshotCache,
   getProviderSnapshot,
+  peekSnapshotEntry,
 } from '../provider-snapshot.js';
 import { loadProviderConfig } from '../provider-config-registry.js';
 
@@ -293,6 +294,49 @@ describe('getProviderSnapshot', () => {
     expect(boocode?.installed).toBe(true);
   });
 
+  it('config models REPLACE the claude static list; additionalModels merge (+ thinking)', async () => {
+    loadConfigFixture({
+      claude: {
+        models: [{ id: 'claude-opus-4-8', label: 'Opus 4.8' }],
+        additionalModels: [{ id: 'sonnet', label: 'Sonnet (latest)' }],
+      },
+    });
+
+    const sql = mockSql([
+      {
+        name: 'claude',
+        install_path: '/usr/bin/claude',
+        supports_acp: false,
+        models: [{ id: 'old-static', label: 'Old' }],
+        label: 'Claude Code',
+        transport: 'pty',
+        last_probed_at: new Date().toISOString(),
+      },
+    ]);
+
+    const entries = await getProviderSnapshot(sql, config, '/tmp/project', true);
+    const claude = entries.find((e) => e.name === 'claude');
+    const ids = claude!.models.map((m) => m.id);
+
+    expect(ids).toContain('claude-opus-4-8'); // config models replaced the DB/static list
+    expect(ids).toContain('sonnet'); // additionalModels merged on top
+    expect(ids).not.toContain('old-static'); // replaced, not appended
+    // thinking options still attach to the config-provided models
+    expect(claude!.models.find((m) => m.id === 'claude-opus-4-8')?.thinkingOptions?.length).toBeGreaterThan(0);
+  });
+
+  it('peekSnapshotEntry returns a cached entry (read-only) and undefined when cold/unknown', async () => {
+    loadConfigFixture({});
+    // Cold cache → undefined (no build triggered).
+    expect(peekSnapshotEntry('boocode', '/tmp/peek')).toBeUndefined();
+
+    const sql = mockSql([]);
+    await getProviderSnapshot(sql, config, '/tmp/peek', true);
+
+    expect(peekSnapshotEntry('boocode', '/tmp/peek')?.name).toBe('boocode');
+    expect(peekSnapshotEntry('does-not-exist', '/tmp/peek')).toBeUndefined();
+  });
+
   it('2.7 warm cache: a second snapshot within the warm window spawns ZERO probes', async () => {
     loadConfigFixture({});
     mockProbe.mockResolvedValue({

apps/coder/src/services/acp-client-fs.ts

@@ -1,5 +1,25 @@
 import { promises as fs } from 'node:fs';
-import { dirname, isAbsolute, join, resolve } from 'node:path';
+import { dirname, isAbsolute, resolve, sep } from 'node:path';
+
+/**
+ * Resolve an ACP-supplied path against the agent worktree and reject anything
+ * that escapes it. Mirrors `write_guard.ts`'s check: `resolve()` to normalize
+ * `../` segments, then a **separator-bounded** prefix test — a bare
+ * `startsWith(root)` wrongly admits a sibling dir like `<root>-evil/...`.
+ *
+ * No realpath (consistent with `write_guard.ts`: the target may not exist yet on
+ * write). This is a containment guard for the ACP fs bridge, not a hard trust
+ * boundary — the agent process already runs with host FS access; symlink-swap
+ * hardening (`O_NOFOLLOW`/realpath) is out of scope here.
+ */
+function resolveInWorktree(worktreePath: string, filePath: string): string {
+  const root = resolve(worktreePath);
+  const absolute = isAbsolute(filePath) ? resolve(filePath) : resolve(root, filePath);
+  if (absolute !== root && !absolute.startsWith(root + sep)) {
+    throw new Error(`path escapes worktree: ${filePath}`);
+  }
+  return absolute;
+}
 
 /** Resolve an ACP path against the agent worktree and read a slice of lines. */
 export async function readWorktreeTextFile(
@@ -8,10 +28,7 @@ export async function readWorktreeTextFile(
   line?: number | null,
   limit?: number | null,
 ): Promise<string> {
-  const absolute = isAbsolute(filePath) ? filePath : resolve(worktreePath, filePath);
+  const absolute = resolveInWorktree(worktreePath, filePath);
-  if (!absolute.startsWith(resolve(worktreePath))) {
-    throw new Error(`path escapes worktree: ${filePath}`);
-  }
   const raw = await fs.readFile(absolute, 'utf8');
   if (!line && !limit) return raw;
   const lines = raw.split(/\r?\n/);
@@ -26,10 +43,7 @@ export async function writeWorktreeTextFile(
   filePath: string,
   content: string,
 ): Promise<void> {
-  const absolute = isAbsolute(filePath) ? filePath : resolve(worktreePath, filePath);
+  const absolute = resolveInWorktree(worktreePath, filePath);
-  if (!absolute.startsWith(resolve(worktreePath))) {
-    throw new Error(`path escapes worktree: ${filePath}`);
-  }
   await fs.mkdir(dirname(absolute), { recursive: true });
   await fs.writeFile(absolute, content, 'utf8');
 }

apps/coder/src/services/acp-probe.ts

@@ -130,6 +130,17 @@ export async function probeAcpProvider(
     });
 
     const session = await connection.newSession({ cwd, mcpServers: [] });
+    // available_commands_update is an async session notification opencode sends
+    // shortly AFTER newSession resolves — reading probedCommands synchronously
+    // here races it and captures nothing. Wait briefly for the first batch, then
+    // a short settle for any stragglers (capped well under PROBE_TIMEOUT_MS).
+    const deadline = Date.now() + 3_000;
+    while (probedCommands.length === 0 && Date.now() < deadline) {
+      await new Promise((r) => setTimeout(r, 150));
+    }
+    if (probedCommands.length > 0) {
+      await new Promise((r) => setTimeout(r, 300));
+    }
     const result = parseSessionResponse(session, agent);
     result.commands = probedCommands;
     await connection.closeSession({ sessionId: session.sessionId }).catch(() => {});

apps/coder/src/services/agent-backend.ts

@@ -0,0 +1,85 @@
+/**
+ * v2.6 — AgentBackend abstraction (Phase 0 scaffold; types only, zero runtime logic).
+ *
+ * The core abstraction for persistent agent sessions. Two implementations land
+ * later: `OpenCodeServerBackend` (Phase 1, opencode HTTP server) and
+ * `WarmAcpBackend` (Phase 2, long-lived ACP process). Backends emit
+ * transport-agnostic `AgentEvent`s; the dispatcher maps them to WS frames.
+ *
+ * Nothing imports this file yet — it must compile standalone.
+ * Spec: openspec/changes/v2-6-persistent-agent-sessions/design.md §2.
+ */
+import type { AcpToolSnapshot } from './acp-tool-snapshot.js';
+import type { AgentCommand } from './provider-types.js';
+
+/** Backend transport kind. Mirrors `agent_sessions.backend` CHECK in schema.sql. */
+export type AgentBackendKind = 'opencode_server' | 'acp_warm';
+
+/**
+ * Normalized, transport-agnostic events a backend emits during a turn (§2).
+ * Derived from acp-dispatch's session-update handling, but WITHOUT the WS
+ * envelope (message_id/chat_id) — the dispatcher owns frame mapping.
+ *
+ * `tool_call` vs `tool_update` are kept distinct on purpose: acp-dispatch
+ * currently merges both into one snapshot frame, but opencode's SSE
+ * distinguishes tool-start from tool-result, so the contract carries both.
+ * `commands` mirrors the ACP `available_commands_update` path (v2.5.10).
+ */
+export type AgentEvent =
+  | { type: 'text'; text: string }
+  | { type: 'reasoning'; text: string }
+  | { type: 'tool_call'; toolCall: AcpToolSnapshot }
+  | { type: 'tool_update'; toolCall: AcpToolSnapshot }
+  | { type: 'commands'; commands: AgentCommand[] };
+
+/** Params to establish (or look up) a backend session (§2). */
+export interface EnsureSessionOpts {
+  agent: string;
+  /** Resolved model id. */
+  model: string;
+  /** Shared per-session worktree (one per `sessions.id`, not per pane). */
+  worktreePath: string;
+  projectId: string;
+}
+
+/** Opaque handle to a live backend session, persisted to `agent_sessions` (§2). */
+export interface AgentSessionHandle {
+  sessionId: string;
+  agent: string;
+  backend: AgentBackendKind;
+  /** Provider's own session id (resume token); null until the backend assigns one. */
+  agentSessionId: string | null;
+  /** opencode HTTP server port; null for ACP backends. */
+  serverPort: number | null;
+}
+
+/** Per-turn context passed to `prompt` (§2). */
+export interface PromptCtx {
+  worktreePath: string;
+  model: string;
+  signal: AbortSignal;
+  onEvent: (e: AgentEvent) => void;
+}
+
+/** Result of a completed turn (§2). Diff/persist happen outside the backend. */
+export interface TurnResult {
+  ok: boolean;
+  error?: string;
+}
+
+/**
+ * The core backend abstraction (§2). Implementations: OpenCodeServerBackend
+ * (Phase 1), WarmAcpBackend (Phase 2).
+ */
+export interface AgentBackend {
+  /** Lazy: spawn server / warm process if not already up for this (session, agent). §2 */
+  ensureSession(sessionId: string, opts: EnsureSessionOpts): Promise<AgentSessionHandle>;
+  /** Send a prompt; stream events via ctx.onEvent; resolves when the turn completes. §2 */
+  prompt(handle: AgentSessionHandle, input: string, ctx: PromptCtx): Promise<TurnResult>;
+  /** Graceful teardown of one session (session close or idle timeout). §2 */
+  closeSession(handle: AgentSessionHandle): Promise<void>;
+  /** Full teardown — kills all spawned servers/processes. §2 */
+  dispose(): Promise<void>;
+  /** Liveness for health endpoint + dispatcher fallback decision. §2 */
+  health(): 'up' | 'down';
+}

apps/coder/src/services/agent-pool.ts

@@ -0,0 +1,44 @@
+/**
+ * v2.6 — AgentPool (Phase 0 scaffold).
+ *
+ * Lazy get-or-create registry of `AgentBackend` instances keyed by
+ * `${sessionId}:${agent}`. Phase 0 ships the skeleton only: an in-memory Map,
+ * lookup / register / health, and clean disposal wired to the server's onClose.
+ * Spawning lands in Phase 1/2; nothing populates the map yet.
+ *
+ * Spec: openspec/changes/v2-6-persistent-agent-sessions/design.md §2.
+ */
+import type { AgentBackend } from './agent-backend.js';
+
+export class AgentPool {
+  private readonly backends = new Map<string, AgentBackend>();
+
+  private key(sessionId: string, agent: string): string {
+    return `${sessionId}:${agent}`;
+  }
+
+  /** Map lookup only. Spawning is Phase 1/2 — never creates here. */
+  get(sessionId: string, agent: string): AgentBackend | undefined {
+    return this.backends.get(this.key(sessionId, agent));
+  }
+
+  /** Store a backend instance for this (session, agent). */
+  register(sessionId: string, agent: string, backend: AgentBackend): void {
+    this.backends.set(this.key(sessionId, agent), backend);
+  }
+
+  /** Summary for the health endpoint. */
+  health(): { size: number } {
+    return { size: this.backends.size };
+  }
+
+  /** Dispose every backend and clear the map. Tolerates throwing backends. */
+  async dispose(): Promise<void> {
+    const entries = [...this.backends.values()];
+    this.backends.clear();
+    await Promise.allSettled(entries.map((b) => b.dispose()));
+  }
+}
+
+/** Single shared instance — referenced only by the server's onClose hook in Phase 0. */
+export const agentPool = new AgentPool();

apps/coder/src/services/agent-probe.ts

@@ -4,7 +4,7 @@ import { exec as execCb, execFile as execFileCb } from 'node:child_process';
 import { promisify } from 'node:util';
 import { PROVIDERS_BY_NAME } from './provider-registry.js';
 import { resolveAcpProbeBinaries } from './acp-spawn.js';
-import { clearProviderSnapshotCache } from './provider-snapshot.js';
+import { clearProviderSnapshotCache, fetchLlamaSwapModels, prefixLlamaSwapModels } from './provider-snapshot.js';
 import { readQwenSettingsModels } from './qwen-settings.js';
 import { loadConfig } from '../config.js';
 import { loadProviderConfig } from './provider-config-registry.js';
@@ -117,6 +117,15 @@ export async function probeAgents(sql: Sql, log: FastifyBaseLogger): Promise<voi
         if (agentName === 'qwen') {
           models = await readQwenSettingsModels();
         }
+        if (providerDef?.mergeLlamaSwap) {
+          try {
+            const config = loadConfig();
+            const llamaModels = prefixLlamaSwapModels(await fetchLlamaSwapModels(config));
+            models = [...models, ...llamaModels];
+          } catch (err) {
+            log.warn({ agent: agentName, err: err instanceof Error ? err.message : String(err) }, 'agent-probe: llama-swap model fetch failed (non-fatal)');
+          }
+        }
       }
 
       const label = resolved.configLabel ?? resolved.label;

apps/coder/src/services/backends/opencode-server.ts

@@ -0,0 +1,748 @@
+/**
+ * v2.6 Phase 1 — OpenCodeServerBackend.
+ *
+ * Warm, multi-turn backend for the `opencode` agent. One `opencode serve` HTTP
+ * server per BooCoder process; one opencode session per BooCode session (resumed
+ * on switch-back); a single SSE read loop demuxes all sessions' events.
+ *
+ * Implements the Phase 0 `AgentBackend` interface. Emits transport-agnostic
+ * `AgentEvent`s — the dispatcher (Phase 1.7, NOT wired in this batch) maps them
+ * to WS frames. No dispatcher/route references this file yet.
+ *
+ * Spec: openspec/changes/v2-6-persistent-agent-sessions/design.md §2 / §2a.
+ * SDK shapes verified by direct read of @opencode-ai/sdk@1.15.12 dist .d.ts:
+ *   - client methods take FLATTENED params (sessionID/directory/body all inline),
+ *     not {path,query,body}. create→{directory}, promptAsync→{sessionID,directory,
+ *     parts,model}, abort→{sessionID,directory}. model is {providerID,modelID}.
+ *   - client.event() resolves to { stream: AsyncGenerator<GlobalEvent> }; the
+ *     real event is chunk.payload (discriminate on chunk.payload.type).
+ *   - promptAsync is fire-and-forget (204); the turn completes via a
+ *     'session.idle' event for that opencode session id.
+ */
+import { spawn, type ChildProcess } from 'node:child_process';
+import { createHash } from 'node:crypto';
+import { createServer } from 'node:net';
+import type { FastifyBaseLogger } from 'fastify';
+import {
+  createOpencodeClient,
+  type OpencodeClient,
+  type Event,
+  type Part,
+  type ToolPart,
+  type ToolState,
+  type AssistantMessage,
+} from '@opencode-ai/sdk/v2/client';
+import type { ToolCallStatus } from '@agentclientprotocol/sdk';
+import type { Sql } from '../../db.js';
+import type { AcpToolSnapshot } from '../acp-tool-snapshot.js';
+import type {
+  AgentBackend,
+  AgentEvent,
+  AgentSessionHandle,
+  EnsureSessionOpts,
+  PromptCtx,
+  TurnResult,
+} from '../agent-backend.js';
+
+const READY_TIMEOUT_MS = 30_000;
+const SSE_RECONNECT_DELAY_MS = 1_000;
+/**
+ * No-activity backstop for an in-flight turn. opencode streams reasoning/text/tool
+ * deltas continuously while working, so "zero events for this long" means the turn
+ * is wedged or its terminal event (session.idle) was lost (see the reconnect race
+ * below). Generous so a legitimately slow turn never trips it.
+ */
+const TURN_INACTIVITY_MS = 180_000;
+
+/** One in-flight turn's emitter + completion settler. */
+interface TurnState {
+  onEvent: (e: AgentEvent) => void;
+  settle: (r: TurnResult) => void;
+}
+
+/** Per-(opencode session) demux state. dedup sets scoped here, cleared per turn. */
+interface SessionState {
+  boocodeSessionId: string;
+  agentSessionId: string;
+  /** Worktree directory for SDK `directory` routing; refreshed each turn from ctx. */
+  worktreePath: string;
+  /** dedup gate: `${type}:${id}` added on delta, deleted-and-tested on updated. Cleared at turn end. */
+  streamedPartKeys: Set<string>;
+  /** partID → 'text' | 'reasoning', so a delta with a non-'reasoning' field is still classed right. Cleared at turn end. */
+  partTypeById: Map<string, string>;
+  activeTurn: TurnState | null;
+  /** Inactivity backstop timer for the active turn; null when no turn in flight. */
+  watchdog: ReturnType<typeof setTimeout> | null;
+}
+
+export interface OpenCodeServerBackendDeps {
+  sql: Sql;
+  log: FastifyBaseLogger;
+  /** Absolute path to the opencode binary (resolved from available_agents at wiring time, Phase 1.7). */
+  opencodeBinary: string;
+}
+
+export class OpenCodeServerBackend implements AgentBackend {
+  readonly backend = 'opencode_server' as const;
+
+  private readonly sql: Sql;
+  private readonly log: FastifyBaseLogger;
+  private readonly opencodeBinary: string;
+
+  private child: ChildProcess | null = null;
+  private client: OpencodeClient | null = null;
+  private port: number | null = null;
+  private up = false;
+  private serverStarting: Promise<void> | null = null;
+  private sseRunning = false;
+
+  /** opencode session id → demux state. Maintained by ensureSession; read by the SSE loop. */
+  private readonly byOpencodeId = new Map<string, SessionState>();
+
+  constructor(deps: OpenCodeServerBackendDeps) {
+    this.sql = deps.sql;
+    this.log = deps.log;
+    this.opencodeBinary = deps.opencodeBinary;
+  }
+
+  /** §2: liveness for the health endpoint + dispatcher fallback decision. */
+  health(): 'up' | 'down' {
+    return this.up ? 'up' : 'down';
+  }
+
+  // ─── Server lifecycle (1.2: spawn once + client + ready) ─────────────────────
+
+  /** Lazy: start the single server on first use. Idempotent — one server per backend. */
+  private ensureServer(): Promise<void> {
+    if (!this.serverStarting) this.serverStarting = this.startServer();
+    return this.serverStarting;
+  }
+
+  private async startServer(): Promise<void> {
+    const port = await freePort();
+
+    // Phase 1: run unsecured on loopback (opencode's documented default — serve.ts
+    // only WARNS when OPENCODE_SERVER_PASSWORD is unset). The real boundary is the
+    // 127.0.0.1 bind. Defense-in-depth basic-auth is deferred: the hey-api client's
+    // auth wiring + opencode's exact scheme must be confirmed against a live server
+    // first, else every request 401s. Recon explicitly said "do NOT block on it".
+    const child = spawn(this.opencodeBinary, ['serve', '--hostname', '127.0.0.1', '--port', String(port)], {
+      stdio: ['ignore', 'pipe', 'pipe'],
+      env: { ...process.env },
+    });
+    this.child = child;
+    this.port = port;
+
+    // Child lifetime is the backend's (the pool's), NOT a request's. We never tie
+    // it to a per-turn abort signal. On unexpected exit we mark down + log; crash
+    // recovery is Phase 3.
+    child.on('exit', (code, signal) => {
+      this.up = false;
+      this.log.warn({ code, signal, port }, 'opencode-server: child exited (recovery is Phase 3)');
+    });
+
+    await waitForReady(child, READY_TIMEOUT_MS);
+
+    this.client = createOpencodeClient({ baseUrl: `http://127.0.0.1:${port}` });
+    this.up = true;
+    this.log.info({ port }, 'opencode-server: ready');
+  }
+
+  // ─── SSE read loop + demux + translate (1.3) + dedup (1.4) ───────────────────
+
+  /** Per-directory SSE subscription. opencode scopes events by directory (defaults
+   *  to process.cwd if omitted) — so we must subscribe with the same directory used
+   *  to create the session. Called from ensureSession; reconnects while up. */
+  private startEventLoop(directory: string): void {
+    if (this.sseRunning) return;
+    this.sseRunning = true;
+    this.sseDirectory = directory;
+    void this.runEventLoop(directory);
+  }
+
+  private sseDirectory: string | null = null;
+
+  private async runEventLoop(directory: string): Promise<void> {
+    while (this.up && this.client) {
+      try {
+        const sub = await this.client.event.subscribe({ directory });
+        for await (const ev of sub.stream) {
+          this.dispatchEvent(ev);
+        }
+        if (this.up) {
+          await this.reconcileInFlight();
+          await sleep(SSE_RECONNECT_DELAY_MS);
+        }
+      } catch (err) {
+        if (!this.up) break;
+        this.log.warn({ err: errMsg(err) }, 'opencode-server: event loop error; reconnecting');
+        await this.reconcileInFlight();
+        await sleep(SSE_RECONNECT_DELAY_MS);
+      }
+    }
+    this.sseRunning = false;
+  }
+
+  /** Demux one event to the owning session's active turn. Unknown/between-turns → drop. */
+  private dispatchEvent(ev: Event): void {
+    switch (ev.type) {
+      // ─── session.next.* — live streaming events (the primary path) ─────────
+      case 'session.next.text.delta': {
+        const p = ev.properties;
+        const st = this.byOpencodeId.get(p.sessionID);
+        if (!st?.activeTurn) return;
+        this.bumpActivity(st);
+        const cleaned = stripDcpTags(p.delta);
+        if (cleaned) st.activeTurn.onEvent({ type: 'text', text: cleaned });
+        return;
+      }
+      case 'session.next.reasoning.delta': {
+        const p = ev.properties;
+        const st = this.byOpencodeId.get(p.sessionID);
+        if (!st?.activeTurn) return;
+        this.bumpActivity(st);
+        st.activeTurn.onEvent({ type: 'reasoning', text: p.delta });
+        return;
+      }
+      case 'session.next.tool.called': {
+        const p = ev.properties;
+        const st = this.byOpencodeId.get(p.sessionID);
+        if (!st?.activeTurn) return;
+        this.bumpActivity(st);
+        const snap: AcpToolSnapshot = {
+          toolCallId: p.callID,
+          title: p.tool,
+          kind: null,
+          status: 'in_progress',
+          rawInput: p.input,
+          rawOutput: undefined,
+        };
+        st.activeTurn.onEvent({ type: 'tool_call', toolCall: snap });
+        return;
+      }
+      case 'session.next.tool.success': {
+        const p = ev.properties;
+        const st = this.byOpencodeId.get(p.sessionID);
+        if (!st?.activeTurn) return;
+        this.bumpActivity(st);
+        const output = p.content?.map((c) => ('text' in c ? (c as { text: string }).text : '')).join('') ?? '';
+        const snap: AcpToolSnapshot = {
+          toolCallId: p.callID,
+          title: p.callID,
+          kind: null,
+          status: 'completed',
+          rawInput: undefined,
+          rawOutput: output,
+        };
+        st.activeTurn.onEvent({ type: 'tool_update', toolCall: snap });
+        return;
+      }
+      case 'session.next.tool.failed': {
+        const p = ev.properties;
+        const st = this.byOpencodeId.get(p.sessionID);
+        if (!st?.activeTurn) return;
+        this.bumpActivity(st);
+        const snap: AcpToolSnapshot = {
+          toolCallId: p.callID,
+          title: p.callID,
+          kind: null,
+          status: 'failed',
+          rawInput: undefined,
+          rawOutput: errToString(p.error),
+        };
+        st.activeTurn.onEvent({ type: 'tool_update', toolCall: snap });
+        return;
+      }
+      // ─── message.part.* — terminal/post-hoc events (dedup gate) ────────────
+      case 'message.part.delta': {
+        const p = ev.properties;
+        const st = this.byOpencodeId.get(p.sessionID);
+        if (!st?.activeTurn) return;
+        this.bumpActivity(st);
+        const isReasoning = p.field === 'reasoning' || st.partTypeById.get(p.partID) === 'reasoning';
+        if (isReasoning) {
+          st.streamedPartKeys.add(`reasoning:${p.partID}`);
+          st.activeTurn.onEvent({ type: 'reasoning', text: p.delta });
+        } else if (p.field === 'text') {
+          st.streamedPartKeys.add(`text:${p.partID}`);
+          const cleaned = stripDcpTags(p.delta);
+          if (cleaned) st.activeTurn.onEvent({ type: 'text', text: cleaned });
+        }
+        return;
+      }
+      case 'message.part.updated': {
+        const part = ev.properties.part;
+        const st = this.byOpencodeId.get(part.sessionID);
+        if (!st?.activeTurn) return;
+        this.bumpActivity(st);
+        this.handleUpdatedPart(part, st);
+        return;
+      }
+      // ─── lifecycle ─────────────────────────────────────────────────────────
+      case 'session.idle': {
+        this.byOpencodeId.get(ev.properties.sessionID)?.activeTurn?.settle({ ok: true });
+        return;
+      }
+      case 'session.error': {
+        const sid = ev.properties.sessionID;
+        if (!sid) return;
+        this.byOpencodeId.get(sid)?.activeTurn?.settle({ ok: false, error: errToString(ev.properties.error) });
+        return;
+      }
+      default:
+        return;
+    }
+  }
+
+  /** Terminal part: dedup gate for text/reasoning; tool parts → tool_call/tool_update. */
+  private handleUpdatedPart(part: Part, st: SessionState): void {
+    const turn = st.activeTurn;
+    if (!turn) return;
+
+    if (part.type === 'text' || part.type === 'reasoning') {
+      st.partTypeById.set(part.id, part.type);
+      const key = resolvePartDedupeKey(part, part.type);
+      if (key && st.streamedPartKeys.delete(key)) return; // already streamed via delta
+      const raw = part.text ?? '';
+      const text = part.type === 'text' ? stripDcpTags(raw) : raw;
+      if (text && part.time?.end != null) {
+        turn.onEvent({ type: part.type, text });
+      }
+      return;
+    }
+
+    if (part.type === 'tool') {
+      const snap = toolPartToSnapshot(part);
+      const status = part.state?.status;
+      // tool_call on start (pending/running), tool_update on terminal (completed/error).
+      // The current ACP path merges both into one frame; the contract keeps them
+      // distinct because opencode's SSE distinguishes start from result.
+      const event: AgentEvent =
+        status === 'completed' || status === 'error'
+          ? { type: 'tool_update', toolCall: snap }
+          : { type: 'tool_call', toolCall: snap };
+      turn.onEvent(event);
+      return;
+    }
+    // NOTE: opencode's SSE payload union carries no available-commands event, so the
+    // AgentEvent 'commands' arm is intentionally never emitted here (1.3).
+  }
+
+  // ─── turn-completion resilience (watchdog + reconnect reconcile) ─────────────
+
+  /** Reset the inactivity backstop on any event routed to a session's active turn. */
+  private bumpActivity(st: SessionState): void {
+    if (!st.activeTurn) return;
+    if (st.watchdog) clearTimeout(st.watchdog);
+    st.watchdog = setTimeout(() => {
+      void this.onTurnStall(st);
+    }, TURN_INACTIVITY_MS);
+    st.watchdog.unref?.();
+  }
+
+  /** Watchdog fired: reconcile once; if the server says still-running we can't tell, so fail closed.
+   *  Also mark the agent_sessions row crashed so a stale session isn't resumed next turn. */
+  private async onTurnStall(st: SessionState): Promise<void> {
+    const settled = await this.reconcile(st);
+    if (!settled) {
+      this.log.warn({ agentSessionId: st.agentSessionId }, 'opencode-server: turn stalled (no activity), failing + marking crashed');
+      await this.sql`
+        UPDATE agent_sessions SET status = 'crashed'
+        WHERE agent_session_id = ${st.agentSessionId}
+      `.catch(() => {});
+      st.activeTurn?.settle({ ok: false, error: 'turn timed out (no activity)' });
+    }
+  }
+
+  /** Reconcile every in-flight turn against the server (called after an SSE drop). */
+  private async reconcileInFlight(): Promise<void> {
+    const states = [...this.byOpencodeId.values()].filter((s) => s.activeTurn);
+    if (states.length === 0) return;
+    await Promise.allSettled(states.map((s) => this.reconcile(s)));
+  }
+
+  /**
+   * Ask the server whether this session's turn already finished — recovers a
+   * session.idle/error lost during an SSE gap. Returns true if it settled the turn.
+   * Inconclusive (still running / call failed) → false; the watchdog covers that.
+   */
+  private async reconcile(st: SessionState): Promise<boolean> {
+    const turn = st.activeTurn;
+    if (!turn || !this.client) return false;
+    try {
+      const res = await this.client.session.messages({
+        sessionID: st.agentSessionId,
+        directory: st.worktreePath,
+      });
+      if (res.error || !res.data) return false;
+      let lastAssistant: AssistantMessage | undefined;
+      for (let i = res.data.length - 1; i >= 0; i--) {
+        const info = res.data[i]!.info;
+        if (info.role === 'assistant') {
+          lastAssistant = info;
+          break;
+        }
+      }
+      if (!lastAssistant) return false;
+      if (lastAssistant.error != null) {
+        turn.settle({ ok: false, error: errToString(lastAssistant.error) });
+        return true;
+      }
+      if (lastAssistant.time.completed != null) {
+        turn.settle({ ok: true });
+        return true;
+      }
+      return false; // still running — the live stream will deliver session.idle
+    } catch {
+      return false; // inconclusive — watchdog backstop covers it
+    }
+  }
+
+  // ─── ensureSession: create-or-resume against agent_sessions (1.5) ────────────
+
+  async ensureSession(sessionId: string, opts: EnsureSessionOpts): Promise<AgentSessionHandle> {
+    await this.ensureServer();
+    if (!this.client) throw new Error('opencode-server: client not ready after ensureServer');
+
+    const configHash = sessionConfigHash(opts.model);
+    const [row] = await this.sql<{ agent_session_id: string | null; status: string; config_hash: string | null }[]>`
+      SELECT agent_session_id, status, config_hash FROM agent_sessions
+      WHERE session_id = ${sessionId} AND agent = ${opts.agent}
+    `;
+    let agentSessionId = row?.agent_session_id ?? null;
+
+    // Don't resume crashed sessions or sessions whose config drifted (model change).
+    const shouldResume = agentSessionId
+      && row!.status !== 'crashed'
+      && (row!.config_hash == null || row!.config_hash === configHash);
+
+    if (!shouldResume) {
+      if (agentSessionId) {
+        this.log.info({ sessionId, oldStatus: row!.status, hashMatch: row!.config_hash === configHash },
+          'opencode-server: not resuming stale session, creating fresh');
+        this.byOpencodeId.delete(agentSessionId);
+      }
+      const created = await this.client.session.create({ directory: opts.worktreePath });
+      if (created.error || !created.data) {
+        throw new Error(`opencode-server: session.create failed: ${errToString(created.error)}`);
+      }
+      agentSessionId = created.data.id;
+      await this.sql`
+        INSERT INTO agent_sessions
+          (session_id, agent, backend, agent_session_id, server_port, status, last_active_at, config_hash)
+        VALUES
+          (${sessionId}, ${opts.agent}, 'opencode_server', ${agentSessionId}, ${this.port}, 'active', clock_timestamp(), ${configHash})
+        ON CONFLICT (session_id, agent) DO UPDATE SET
+          backend = 'opencode_server',
+          agent_session_id = EXCLUDED.agent_session_id,
+          server_port = EXCLUDED.server_port,
+          status = 'active',
+          last_active_at = clock_timestamp(),
+          config_hash = EXCLUDED.config_hash
+      `;
+    } else {
+      await this.sql`
+        UPDATE agent_sessions
+        SET status = 'active', last_active_at = clock_timestamp(), server_port = ${this.port}, config_hash = ${configHash}
+        WHERE session_id = ${sessionId} AND agent = ${opts.agent}
+      `;
+    }
+
+    // Both branches above guarantee agentSessionId is non-null.
+    const ocSessionId = agentSessionId!;
+
+    // Start (or re-start) the SSE event loop scoped to this session's directory.
+    // opencode scopes events by the `directory` query param; without it events
+    // default to the server's CWD which doesn't match our worktree paths.
+    //
+    // KNOWN Phase 1 LIMITATION: one SSE stream at a time, scoped to a single
+    // directory. Under 1.9 concurrency, if two opencode sessions use different
+    // worktree directories simultaneously, re-subscribing for the second drops
+    // the first session's events (the watchdog backstop prevents a full hang,
+    // but streamed content is lost). Phase 2 should move to per-session SSE
+    // subscriptions or a directory-agnostic event path.
+    if (!this.sseRunning || this.sseDirectory !== opts.worktreePath) {
+      if (this.sseRunning && this.sseDirectory && this.sseDirectory !== opts.worktreePath) {
+        this.log.warn(
+          { prev: this.sseDirectory, next: opts.worktreePath },
+          'opencode-server: SSE directory changed — concurrent sessions will lose events from the previous directory',
+        );
+      }
+      this.sseRunning = false;
+      this.startEventLoop(opts.worktreePath);
+    }
+
+    // Register / refresh the demux entry the SSE loop keys on. Preserve an existing
+    // entry (and any in-flight turn) — just refresh the routing fields.
+    const existing = this.byOpencodeId.get(ocSessionId);
+    if (existing) {
+      existing.boocodeSessionId = sessionId;
+      existing.worktreePath = opts.worktreePath;
+    } else {
+      this.byOpencodeId.set(ocSessionId, {
+        boocodeSessionId: sessionId,
+        agentSessionId: ocSessionId,
+        worktreePath: opts.worktreePath,
+        streamedPartKeys: new Set(),
+        partTypeById: new Map(),
+        activeTurn: null,
+        watchdog: null,
+      });
+    }
+
+    return {
+      sessionId,
+      agent: opts.agent,
+      backend: 'opencode_server',
+      agentSessionId: ocSessionId,
+      serverPort: this.port,
+    };
+  }
+
+  // ─── prompt: send one turn (1.6) ─────────────────────────────────────────────
+
+  async prompt(handle: AgentSessionHandle, input: string, ctx: PromptCtx): Promise<TurnResult> {
+    if (!this.client) throw new Error('opencode-server: client not ready');
+    const oc = handle.agentSessionId;
+    if (!oc) throw new Error('opencode-server: handle has no agentSessionId');
+
+    let state = this.byOpencodeId.get(oc);
+    if (!state) {
+      state = {
+        boocodeSessionId: handle.sessionId,
+        agentSessionId: oc,
+        worktreePath: ctx.worktreePath,
+        streamedPartKeys: new Set(),
+        partTypeById: new Map(),
+        activeTurn: null,
+        watchdog: null,
+      };
+      this.byOpencodeId.set(oc, state);
+    }
+    const session = state;
+    // Authoritative per-turn directory for SDK routing + reconcile.
+    session.worktreePath = ctx.worktreePath;
+    const client = this.client;
+
+    return await new Promise<TurnResult>((resolve) => {
+      let settled = false;
+      const cleanup = () => {
+        session.activeTurn = null;
+        if (session.watchdog) {
+          clearTimeout(session.watchdog);
+          session.watchdog = null;
+        }
+        session.streamedPartKeys.clear();
+        session.partTypeById.clear();
+        ctx.signal.removeEventListener('abort', onAbort);
+      };
+      const settle = (r: TurnResult) => {
+        if (settled) return;
+        settled = true;
+        cleanup();
+        resolve(r);
+      };
+      const onAbort = () => {
+        // Abort the turn only — never the server.
+        client.session.abort({ sessionID: oc, directory: ctx.worktreePath }).catch(() => {});
+        settle({ ok: false, error: 'aborted' });
+      };
+
+      session.activeTurn = { onEvent: ctx.onEvent, settle };
+      this.bumpActivity(session); // arm the inactivity backstop
+
+      if (ctx.signal.aborted) {
+        onAbort();
+        return;
+      }
+      ctx.signal.addEventListener('abort', onAbort, { once: true });
+
+      const model = parseModel(ctx.model);
+      client.session
+        .promptAsync({
+          sessionID: oc,
+          directory: ctx.worktreePath,
+          parts: [{ type: 'text', text: input }],
+          ...(model ? { model } : {}),
+        })
+        .then((res) => {
+          // promptAsync is fire-and-forget (204); the turn completes via session.idle.
+          // Only a submission error settles here.
+          if (res.error) settle({ ok: false, error: errToString(res.error) });
+        })
+        .catch((err) => settle({ ok: false, error: errMsg(err) }));
+    });
+  }
+
+  // ─── teardown ────────────────────────────────────────────────────────────────
+
+  async closeSession(handle: AgentSessionHandle): Promise<void> {
+    if (handle.agentSessionId) this.byOpencodeId.delete(handle.agentSessionId);
+    await this.sql`
+      UPDATE agent_sessions SET status = 'closed'
+      WHERE session_id = ${handle.sessionId} AND agent = ${handle.agent}
+    `.catch(() => {});
+  }
+
+  async dispose(): Promise<void> {
+    this.up = false;
+    const child = this.child;
+    this.child = null;
+    this.client = null;
+    this.byOpencodeId.clear();
+    if (child && !child.killed) {
+      child.kill('SIGTERM');
+      const t = setTimeout(() => {
+        if (!child.killed) child.kill('SIGKILL');
+      }, 5_000);
+      t.unref();
+    }
+  }
+}
+
+// ─── helpers ──────────────────────────────────────────────────────────────────
+
+/** BooCoder model string "provider/model" → opencode's structured {providerID, modelID}. */
+function parseModel(model: string | undefined): { providerID: string; modelID: string } | undefined {
+  if (!model || !model.trim()) return undefined;
+  const trimmed = model.trim();
+  const idx = trimmed.indexOf('/');
+  if (idx > 0 && idx < trimmed.length - 1) {
+    return { providerID: trimmed.slice(0, idx), modelID: trimmed.slice(idx + 1) };
+  }
+  // No slash but non-empty → infer llama-swap (the only configured provider).
+  // Guard against bare '/' or trailing/leading slash.
+  if (idx < 0 && trimmed.length > 0) {
+    return { providerID: 'llama-swap', modelID: trimmed };
+  }
+  return undefined;
+}
+
+/** Ported verbatim from Paseo opencode-agent.ts: id → message-id fallback → null. */
+function resolvePartDedupeKey(part: { id: string; messageID: string }, type: string): string | null {
+  if (part.id.trim().length > 0) return `${type}:${part.id}`;
+  if (part.messageID.trim().length > 0) return `${type}:message:${part.messageID}`;
+  return null;
+}
+
+/** opencode ToolPart → ACP-shaped snapshot (reuses the existing persist/render path). */
+function toolPartToSnapshot(part: ToolPart): AcpToolSnapshot {
+  const state = part.state;
+  let rawInput: unknown;
+  let rawOutput: unknown;
+  let title: string | undefined;
+  if (state) {
+    if ('input' in state) rawInput = (state as { input?: unknown }).input;
+    if ('output' in state) rawOutput = (state as { output?: unknown }).output;
+    else if ('error' in state) rawOutput = (state as { error?: unknown }).error;
+    if ('title' in state) title = (state as { title?: string }).title;
+  }
+  return {
+    toolCallId: part.callID,
+    title: title ?? part.tool,
+    kind: null,
+    status: mapToolStatus(state?.status),
+    rawInput,
+    rawOutput,
+  };
+}
+
+function mapToolStatus(s: ToolState['status'] | undefined): ToolCallStatus | null {
+  switch (s) {
+    case 'pending':
+      return 'pending';
+    case 'running':
+      return 'in_progress';
+    case 'completed':
+      return 'completed';
+    case 'error':
+      return 'failed';
+    default:
+      return null;
+  }
+}
+
+/** Bind-probe an ephemeral port on loopback. */
+function freePort(): Promise<number> {
+  return new Promise((resolve, reject) => {
+    const srv = createServer();
+    srv.unref();
+    srv.on('error', reject);
+    srv.listen(0, '127.0.0.1', () => {
+      const addr = srv.address();
+      if (addr && typeof addr === 'object') {
+        const { port } = addr;
+        srv.close(() => resolve(port));
+      } else {
+        srv.close(() => reject(new Error('opencode-server: could not determine a free port')));
+      }
+    });
+  });
+}
+
+/** Resolve when the child prints the ready line; reject on timeout or early exit. */
+function waitForReady(child: ChildProcess, timeoutMs: number): Promise<void> {
+  return new Promise((resolve, reject) => {
+    let done = false;
+    let stderrBuf = '';
+
+    const finish = (err?: Error) => {
+      if (done) return;
+      done = true;
+      clearTimeout(timer);
+      child.stdout?.off('data', onOut);
+      child.stderr?.off('data', onErr);
+      child.off('exit', onExit);
+      if (err) reject(err);
+      else resolve();
+    };
+
+    const onOut = (buf: Buffer) => {
+      if (buf.toString().includes('opencode server listening on')) finish();
+    };
+    const onErr = (buf: Buffer) => {
+      stderrBuf += buf.toString();
+    };
+    const onExit = (code: number | null) =>
+      finish(new Error(`opencode serve exited before ready (code ${code}); stderr: ${stderrBuf.slice(-2000)}`));
+    const timer = setTimeout(
+      () => finish(new Error(`opencode serve not ready in ${timeoutMs}ms; stderr: ${stderrBuf.slice(-2000)}`)),
+      timeoutMs,
+    );
+
+    child.stdout?.on('data', onOut);
+    child.stderr?.on('data', onErr);
+    child.on('exit', onExit);
+  });
+}
+
+function sleep(ms: number): Promise<void> {
+  return new Promise((r) => setTimeout(r, ms));
+}
+
+/** Strip opencode-dcp plugin tags that render as literal text in the UI. */
+function stripDcpTags(s: string): string {
+  return s.replace(/<dcp-message-id>[^<]*<\/dcp-message-id>/g, '');
+}
+
+function errMsg(e: unknown): string {
+  return e instanceof Error ? e.message : String(e);
+}
+
+function errToString(e: unknown): string {
+  if (e == null) return 'unknown error';
+  if (typeof e === 'string') return e;
+  if (e instanceof Error) return e.message;
+  try {
+    return JSON.stringify(e);
+  } catch {
+    return String(e);
+  }
+}
+
+/** Hash of stable config — detects model changes across sessions without
+ *  invalidating on ephemeral state like the random server port (which changes
+ *  every BooCoder restart). */
+function sessionConfigHash(model: string): string {
+  return createHash('sha256').update(`opencode_server|${model}`).digest('hex').slice(0, 16);
+}

apps/coder/src/services/claude-command-discovery.ts

@@ -0,0 +1,108 @@
+/**
+ * v2.5.11: discover Claude Code's real, enabled commands + plugin skills from
+ * disk so the coder slash menu shows them (claude is PTY — no ACP discovery).
+ *
+ * Scope (v1): user-global only — `~/.claude/commands/*.md` plus the enabled
+ * plugins listed in `~/.claude/settings.json:enabledPlugins` (user-scope install
+ * paths from `~/.claude/plugins/.../installed_plugins.json`). Project-local
+ * plugins and `<cwd>/.claude/commands` are deferred. Names are bare.
+ */
+import { readFileSync, readdirSync, existsSync } from 'node:fs';
+import { homedir } from 'node:os';
+import { join } from 'node:path';
+import type { AgentCommand } from './provider-types.js';
+
+/** Minimal frontmatter reader — single-line `key: value` between `---` fences. */
+function frontmatterField(content: string, field: string): string | undefined {
+  const block = content.match(/^---\r?\n([\s\S]*?)\r?\n---/);
+  if (!block?.[1]) return undefined;
+  const m = block[1].match(new RegExp(`^${field}:\\s*(.+)$`, 'm'));
+  return m?.[1]?.trim().replace(/^["']|["']$/g, '') || undefined;
+}
+
+function readCommandDir(dir: string): AgentCommand[] {
+  if (!existsSync(dir)) return [];
+  let files: string[];
+  try {
+    files = readdirSync(dir);
+  } catch {
+    return [];
+  }
+  const out: AgentCommand[] = [];
+  for (const f of files) {
+    if (!f.endsWith('.md')) continue;
+    let description: string | undefined;
+    try {
+      description = frontmatterField(readFileSync(join(dir, f), 'utf8'), 'description');
+    } catch {
+      /* unreadable — still list the command by name */
+    }
+    out.push({ name: f.slice(0, -3), kind: 'command', ...(description ? { description } : {}) });
+  }
+  return out;
+}
+
+function readSkillDir(dir: string): AgentCommand[] {
+  if (!existsSync(dir)) return [];
+  let entries: string[];
+  try {
+    entries = readdirSync(dir);
+  } catch {
+    return [];
+  }
+  const out: AgentCommand[] = [];
+  for (const sub of entries) {
+    const skillMd = join(dir, sub, 'SKILL.md');
+    if (!existsSync(skillMd)) continue;
+    let content: string;
+    try {
+      content = readFileSync(skillMd, 'utf8');
+    } catch {
+      continue;
+    }
+    out.push({
+      name: frontmatterField(content, 'name') ?? sub,
+      kind: 'skill',
+      ...(() => {
+        const d = frontmatterField(content, 'description');
+        return d ? { description: d } : {};
+      })(),
+    });
+  }
+  return out;
+}
+
+export function discoverClaudeCommands(): AgentCommand[] {
+  const root = join(homedir(), '.claude');
+  const out: AgentCommand[] = [];
+
+  // User custom commands.
+  out.push(...readCommandDir(join(root, 'commands')));
+
+  // Enabled plugins (user-scope installs).
+  try {
+    const settings = JSON.parse(readFileSync(join(root, 'settings.json'), 'utf8')) as {
+      enabledPlugins?: Record<string, boolean>;
+    };
+    const installed = JSON.parse(
+      readFileSync(join(root, 'plugins', 'installed_plugins.json'), 'utf8'),
+    ) as { plugins?: Record<string, Array<{ scope?: string; installPath?: string }>> };
+
+    const enabled = settings.enabledPlugins ?? {};
+    const plugins = installed.plugins ?? {};
+    for (const [key, on] of Object.entries(enabled)) {
+      if (!on) continue;
+      const installs = plugins[key] ?? [];
+      const installPath = (installs.find((i) => i.scope === 'user') ?? installs[0])?.installPath;
+      if (!installPath || !existsSync(installPath)) continue;
+      out.push(...readSkillDir(join(installPath, 'skills')));
+      out.push(...readCommandDir(join(installPath, 'commands')));
+    }
+  } catch {
+    /* missing/unreadable plugin config → user commands only */
+  }
+
+  // Dedupe by name (first wins).
+  const seen = new Set<string>();
+  return out.filter((c) => (seen.has(c.name) ? false : (seen.add(c.name), true)));
+}

apps/coder/src/services/dispatcher.ts

@@ -3,13 +3,17 @@ import type { FastifyBaseLogger } from 'fastify';
 import type { Broker } from '@boocode/server/broker';
 import type { WsFrame } from '@boocode/server/ws-frames';
 import type { Config } from '../config.js';
-import { createWorktree, diffWorktree, cleanupWorktree } from './worktrees.js';
+import { createWorktree, diffWorktree, cleanupWorktree, ensureSessionWorktree } from './worktrees.js';
 import { dispatchViaAcp } from './acp-dispatch.js';
 import { getResolvedRegistry } from './provider-config-registry.js';
 import { dispatchViaPty } from './pty-dispatch.js';
 import { clearTaskCommands, setTaskCommands } from './agent-commands-cache.js';
 import { getManifestCommands } from './provider-commands.js';
 import { persistExternalAgentTurn } from './agent-turn-persist.js';
+import { snapshotToWireToolCall, type AcpToolSnapshot } from './acp-tool-snapshot.js';
+import { agentPool } from './agent-pool.js';
+import { OpenCodeServerBackend } from './backends/opencode-server.js';
+import type { AgentBackend, AgentEvent } from './agent-backend.js';
 
 interface InferenceRunner {
   enqueue: (sessionId: string, chatId: string, assistantId: string, user: string) => void;
@@ -35,47 +39,65 @@ export function createDispatcher(deps: Deps): { start(): void; stop(): Promise<v
   const { sql, inference, broker, log, config } = deps;
   let timer: ReturnType<typeof setInterval> | null = null;
   let listener: { unlisten: () => Promise<void> } | null = null;
-  let running = false;
+  let polling = false;
   let stopping = false;
-  let inflightPromise: Promise<void> | null = null;
+  // v2.6 (1.9): per-session in-flight registry replaces the global `running`
+  // boolean. Key = session_id (or `task:<id>` for sessionless tasks). Sessions
+  // without an in-flight turn run concurrently; within a session, strictly one
+  // turn at a time.
+  const inflight = new Map<string, Promise<void>>();
 
   // Shared entry point for both the poll timer and the NOTIFY listener. poll()'s
-  // `running`/`stopping` guard makes this safe to call concurrently — a notify
+  // `polling`/`stopping` guard makes this safe to call concurrently — a notify
-  // arriving mid-task returns immediately and never double-dispatches.
+  // arriving mid-poll returns immediately and never double-dispatches.
   function triggerPoll(reason: string): void {
     poll().catch((err) => {
       log.error({ err, reason }, 'dispatcher: poll error');
     });
   }
 
+  function concurrencyKey(task: { id: string; session_id: string | null }): string {
+    return task.session_id ?? `task:${task.id}`;
+  }
+
   async function poll(): Promise<void> {
-    if (running || stopping) return;
+    // `polling` serializes poll() execution itself (timer + NOTIFY can fire
-
+    // concurrently) so we never double-select a task. It does NOT serialize task
-    // Grab one pending task
+    // execution — that's what `inflight` (keyed per session) governs.
-    const rows = await sql<{
+    if (polling || stopping) return;
-      id: string;
+    polling = true;
-      project_id: string;
+    try {
-      input: string;
+      // Oldest-first; start every pending task whose session isn't already busy.
-      agent: string | null;
+      const rows = await sql<{
-      model: string | null;
+        id: string;
-      mode_id: string | null;
+        project_id: string;
-      thinking_option_id: string | null;
+        input: string;
-      session_id: string | null;
+        agent: string | null;
-    }[]>`
+        model: string | null;
-      SELECT id, project_id, input, agent, model, mode_id, thinking_option_id, session_id
+        mode_id: string | null;
-      FROM tasks
+        thinking_option_id: string | null;
-      WHERE state = 'pending'
+        session_id: string | null;
-      ORDER BY created_at
+      }[]>`
-      LIMIT 1
+        SELECT id, project_id, input, agent, model, mode_id, thinking_option_id, session_id
-    `;
+        FROM tasks
-    if (rows.length === 0) return;
+        WHERE state = 'pending'
-
+        ORDER BY created_at
-    const task = rows[0]!;
+        LIMIT 50
-    running = true;
+      `;
-    inflightPromise = runTask(task).finally(() => {
+      for (const task of rows) {
-      running = false;
+        if (stopping) break;
-      inflightPromise = null;
+        const key = concurrencyKey(task);
-    });
+        if (inflight.has(key)) continue; // this session already has an in-flight turn
+        // Register synchronously (before any await) so a later row in this pass
+        // with the same key is skipped and a concurrent poll can't re-pick it.
+        const p = runTask(task).finally(() => {
+          inflight.delete(key);
+        });
+        inflight.set(key, p);
+      }
+    } finally {
+      polling = false;
+    }
   }
 
   async function runTask(task: {
@@ -96,7 +118,13 @@ export function createDispatcher(deps: Deps): { start(): void; stop(): Promise<v
         SELECT name, supports_acp, install_path FROM available_agents WHERE name = ${task.agent}
       `;
       if (agentRow) {
-        await runExternalAgent(task, agentRow.supports_acp, agentRow.install_path);
+        // v2.6 (1.7): opencode routes to the warm pool backend; every other
+        // external agent keeps the existing one-shot ACP/PTY path untouched.
+        if (task.agent === 'opencode') {
+          await runOpenCodeServerTask(task, agentRow.install_path);
+        } else {
+          await runExternalAgent(task, agentRow.supports_acp, agentRow.install_path);
+        }
         return;
       }
       // Agent specified but not available — fall through to Path A with a warning
@@ -456,6 +484,274 @@ export function createDispatcher(deps: Deps): { start(): void; stop(): Promise<v
     }
   }
 
+  // ─── Path B (opencode): warm OpenCode server backend (v2.6 1.7 + 1.10) ───────
+
+  // OpenCode runs ONE server per BooCoder process, shared across all sessions
+  // (the backend multiplexes sessions internally), so it's pooled under a fixed
+  // key rather than per-session. Warm ACP backends (Phase 2) will be per-session.
+  const OPENCODE_POOL_KEY = '__opencode_server__';
+
+  function getOpenCodeBackend(installPath: string | null): AgentBackend {
+    let backend = agentPool.get(OPENCODE_POOL_KEY, 'opencode');
+    if (!backend) {
+      backend = new OpenCodeServerBackend({ sql, log, opencodeBinary: installPath ?? 'opencode' });
+      agentPool.register(OPENCODE_POOL_KEY, 'opencode', backend);
+    }
+    return backend;
+  }
+
+  async function runOpenCodeServerTask(
+    task: {
+      id: string;
+      project_id: string;
+      input: string;
+      agent: string | null;
+      model: string | null;
+      mode_id: string | null;
+      thinking_option_id: string | null;
+      session_id: string | null;
+    },
+    installPath: string | null,
+  ): Promise<void> {
+    const taskId = task.id;
+    const agent = 'opencode';
+    log.info({ taskId, agent }, 'dispatcher: starting task (path B — opencode server)');
+
+    const [project] = await sql<{ path: string | null }[]>`
+      SELECT path FROM projects WHERE id = ${task.project_id}
+    `;
+    const projectPath = project?.path;
+    if (!projectPath) {
+      await sql`
+        UPDATE tasks
+        SET state = 'failed', ended_at = clock_timestamp(), output_summary = 'Project has no path — cannot create worktree'
+        WHERE id = ${taskId}
+      `;
+      return;
+    }
+
+    const ac = new AbortController();
+
+    try {
+      // execution_path = 'acp' — the schema CHECK has no 'opencode_server' value
+      // (schema is frozen at Phase 0); the warm-vs-one-shot distinction lives in
+      // agent_sessions.backend. Reuse the closest existing value.
+      await sql`
+        UPDATE tasks
+        SET state = 'running', started_at = clock_timestamp(), execution_path = 'acp'
+        WHERE id = ${taskId}
+      `;
+
+      // Resolve session + chat (mirrors runExternalAgent).
+      let sessionId: string;
+      let chatId: string;
+      if (task.session_id) {
+        sessionId = task.session_id;
+        const chats = await sql<{ id: string }[]>`
+          SELECT id FROM chats WHERE session_id = ${sessionId} AND status = 'open' ORDER BY created_at DESC LIMIT 1
+        `;
+        if (chats.length === 0) {
+          const [chat] = await sql<{ id: string }[]>`
+            INSERT INTO chats (session_id, name, status)
+            VALUES (${sessionId}, 'External agent execution', 'open')
+            RETURNING id
+          `;
+          chatId = chat!.id;
+        } else {
+          chatId = chats[0]!.id;
+        }
+      } else {
+        const sessionName = `Task [${agent}]: ${task.input.slice(0, 30)}`;
+        const [session] = await sql<{ id: string }[]>`
+          INSERT INTO sessions (project_id, name, model, status)
+          VALUES (${task.project_id}, ${sessionName}, ${task.model ?? config.DEFAULT_MODEL}, 'open')
+          RETURNING id
+        `;
+        sessionId = session!.id;
+        const [chat] = await sql<{ id: string }[]>`
+          INSERT INTO chats (session_id, name, status)
+          VALUES (${sessionId}, 'External agent execution', 'open')
+          RETURNING id
+        `;
+        chatId = chat!.id;
+        await sql`UPDATE tasks SET session_id = ${sessionId} WHERE id = ${taskId}`;
+      }
+
+      if (!task.session_id) {
+        await sql`
+          INSERT INTO messages (session_id, chat_id, role, content, status, created_at)
+          VALUES (${sessionId}, ${chatId}, 'user', ${task.input}, 'complete', clock_timestamp())
+        `;
+      }
+
+      // Persistent, session-keyed worktree (shared across turns; NOT torn down
+      // per turn — Phase 3 reaps it). Captures base_commit for a stable diff.
+      const { worktreePath, baseCommit } = await ensureSessionWorktree(sql, projectPath, sessionId, {
+        signal: ac.signal,
+      });
+      log.info({ taskId, worktreePath }, 'dispatcher: session worktree ready');
+
+      const [assistantMsg] = await sql<{ id: string }[]>`
+        INSERT INTO messages (session_id, chat_id, role, content, status, created_at)
+        VALUES (${sessionId}, ${chatId}, 'assistant', '', 'streaming', clock_timestamp())
+        RETURNING id
+      `;
+      const assistantId = assistantMsg!.id;
+
+      broker.publishFrame(sessionId, {
+        type: 'message_started',
+        message_id: assistantId,
+        chat_id: chatId,
+        role: 'assistant',
+      } as WsFrame);
+
+      const manifestCommands = getManifestCommands(agent);
+      if (manifestCommands.length > 0) {
+        setTaskCommands(taskId, manifestCommands);
+        broker.publishFrame(sessionId, {
+          type: 'agent_commands',
+          task_id: taskId,
+          session_id: sessionId,
+          commands: manifestCommands,
+        } as WsFrame);
+      }
+
+      // Accumulate the turn's stream for persistence + the final message content.
+      const textChunks: string[] = [];
+      const reasoningChunks: string[] = [];
+      const toolSnaps = new Map<string, AcpToolSnapshot>();
+
+      // Map transport-agnostic AgentEvents → the SAME WS frames the ACP path emits.
+      // This boundary is where message_id/chat_id get attached (the backend never
+      // owns them).
+      const onEvent = (e: AgentEvent): void => {
+        switch (e.type) {
+          case 'text':
+            textChunks.push(e.text);
+            broker.publishFrame(sessionId, {
+              type: 'delta',
+              message_id: assistantId,
+              chat_id: chatId,
+              content: e.text,
+            } as WsFrame);
+            break;
+          case 'reasoning':
+            reasoningChunks.push(e.text);
+            broker.publishFrame(sessionId, {
+              type: 'reasoning_delta',
+              message_id: assistantId,
+              chat_id: chatId,
+              content: e.text,
+            } as WsFrame);
+            break;
+          case 'tool_call':
+          case 'tool_update':
+            toolSnaps.set(e.toolCall.toolCallId, e.toolCall);
+            broker.publishFrame(sessionId, {
+              type: 'tool_call',
+              message_id: assistantId,
+              chat_id: chatId,
+              tool_call: snapshotToWireToolCall(e.toolCall),
+            } as WsFrame);
+            break;
+          case 'commands':
+            // opencode-server doesn't emit these today; ignore if it ever does.
+            break;
+        }
+      };
+
+      // opencode expects provider-prefixed model ids (e.g. 'llama-swap/qwen3.6-35b…').
+      // DEFAULT_MODEL is bare (no prefix) because native inference uses it directly
+      // against llama-swap. Coalesce empty string (frontend sends '' when no models
+      // listed) and prefix bare ids so parseModel always succeeds.
+      const rawModel = (task.model && task.model.trim()) || config.DEFAULT_MODEL;
+      const model = rawModel.includes('/') ? rawModel : `llama-swap/${rawModel}`;
+      const backend = getOpenCodeBackend(installPath);
+      const handle = await backend.ensureSession(sessionId, {
+        agent,
+        model,
+        worktreePath,
+        projectId: task.project_id,
+      });
+      const result = await backend.prompt(handle, task.input, {
+        worktreePath,
+        model,
+        signal: ac.signal,
+        onEvent,
+      });
+
+      const assistantContent = textChunks.join('').slice(0, 50_000);
+      const reasoningText = reasoningChunks.join('').slice(0, 200_000);
+      const outputSummary = (result.ok ? textChunks.join('') : result.error ?? 'opencode turn failed').slice(0, 500);
+
+      await persistExternalAgentTurn(sql, assistantId, [...toolSnaps.values()], reasoningText);
+
+      await sql`
+        UPDATE messages
+        SET content = ${assistantContent}, status = 'complete', finished_at = clock_timestamp()
+        WHERE id = ${assistantId}
+      `;
+      broker.publishFrame(sessionId, {
+        type: 'message_complete',
+        message_id: assistantId,
+        chat_id: chatId,
+      } as WsFrame);
+
+      if (stopping) {
+        await sql`UPDATE tasks SET state = 'cancelled', ended_at = clock_timestamp() WHERE id = ${taskId}`;
+        return; // worktree persists (no cleanup); backend stays warm
+      }
+
+      // 1.10: diff the persistent worktree against its captured baseline and
+      // SUPERSEDE the session's prior pending row (latest-wins, one accumulating
+      // diff) instead of stacking. Stamp agent for DiffPanel attribution.
+      const diff = await diffWorktree(worktreePath, projectPath, {
+        signal: ac.signal,
+        baseRef: baseCommit ?? 'HEAD',
+      });
+      if (diff) {
+        await sql`
+          DELETE FROM pending_changes WHERE session_id = ${sessionId} AND status = 'pending'
+        `;
+        await sql`
+          INSERT INTO pending_changes (session_id, task_id, file_path, operation, diff, agent)
+          VALUES (${sessionId}, ${taskId}, ${projectPath}, 'edit', ${diff}, ${agent})
+        `;
+        log.info({ taskId, diffLength: diff.length }, 'dispatcher: diff superseded prior pending change');
+      } else {
+        log.info({ taskId }, 'dispatcher: no changes detected in session worktree');
+      }
+
+      // NO worktree cleanup — it's persistent (Phase 3 reaps it). Backend stays warm.
+
+      const [extCostRow] = await sql<{ total: number | null }[]>`
+        SELECT SUM(tokens_used)::int AS total
+        FROM messages
+        WHERE session_id = ${sessionId} AND tokens_used IS NOT NULL
+      `;
+      const extCostTokens = extCostRow?.total ?? null;
+
+      const finalState = result.ok ? 'completed' : 'failed';
+      await sql`
+        UPDATE tasks
+        SET state = ${finalState}, ended_at = clock_timestamp(), output_summary = ${outputSummary}, cost_tokens = ${extCostTokens}
+        WHERE id = ${taskId}
+      `;
+      log.info({ taskId, agent, finalState, costTokens: extCostTokens }, 'dispatcher: task finished (opencode server)');
+      clearTaskCommands(taskId);
+    } catch (err) {
+      const errMsg = err instanceof Error ? err.message : String(err);
+      log.error({ taskId, agent, err: errMsg }, 'dispatcher: opencode server error');
+      await sql`
+        UPDATE tasks
+        SET state = 'failed', ended_at = clock_timestamp(), output_summary = ${errMsg.slice(0, 500)}
+        WHERE id = ${taskId}
+      `.catch(() => {});
+      clearTaskCommands(taskId);
+      // No worktree cleanup (persistent); backend stays warm for the next turn.
+    }
+  }
+
   // ─── Helpers ────────────────────────────────────────────────────────────────
 
   async function waitForCompletion(assistantId: string): Promise<string> {
@@ -514,9 +810,9 @@ export function createDispatcher(deps: Deps): { start(): void; stop(): Promise<v
         });
         listener = null;
       }
-      if (inflightPromise) {
+      if (inflight.size > 0) {
-        log.info('dispatcher: waiting for in-flight task');
+        log.info({ count: inflight.size }, 'dispatcher: waiting for in-flight tasks');
-        await inflightPromise;
+        await Promise.allSettled([...inflight.values()]);
       }
       log.info('dispatcher: stopped');
     },

apps/coder/src/services/provider-config-registry.ts

@@ -22,6 +22,10 @@ export interface ResolvedProviderDef extends ProviderDef {
   env: Record<string, string> | undefined;
   configLabel?: string;
   configDescription?: string;
+  /** Config `models` — REPLACES the discovered/static model list when present. */
+  configModels?: Array<{ id: string; label: string }>;
+  /** Config `additionalModels` — MERGED on top of the resolved model list. */
+  configAdditionalModels?: Array<{ id: string; label: string }>;
 }
 
 /**
@@ -61,6 +65,8 @@ export function buildResolvedRegistry(
       env: ov?.env,
       configLabel: ov?.label,
       configDescription: ov?.description,
+      configModels: ov?.models,
+      configAdditionalModels: ov?.additionalModels,
     });
   }
 
@@ -87,6 +93,8 @@ export function buildResolvedRegistry(
       env: ov.env,
       configLabel: ov.label,
       configDescription: ov.description,
+      configModels: ov.models,
+      configAdditionalModels: ov.additionalModels,
     });
   }
 

apps/coder/src/services/provider-config.ts

@@ -29,6 +29,41 @@ export const CoderProvidersFileSchema = z.object({
 export type ProviderOverride = z.infer<typeof ProviderOverrideSchema>;
 export type CoderProvidersFile = z.infer<typeof CoderProvidersFileSchema>;
 
+/**
+ * PATCH body schema (design.md §6.2). A partial providers map where each value
+ * is either a full override object (REPLACES that id's override) or `null`
+ * (DELETES the override → revert to the built-in default). Ids absent from the
+ * patch are left untouched. The route validates the body against this first
+ * (malformed → 422) so a bad shape can never reach the merge/save step.
+ */
+export const ProviderConfigPatchSchema = z.object({
+  providers: z.record(ProviderOverrideSchema.nullable()).default({}),
+});
+
+export type ProviderConfigPatch = z.infer<typeof ProviderConfigPatchSchema>;
+
+/**
+ * Shallow per-id merge (design.md §6.2 / Paseo `patchConfig`). Each key in
+ * `patch.providers` REPLACES that id's override object wholesale (NOT a deep
+ * field merge); a `null` value DELETES the override. Returns a new object —
+ * never mutates `current`. The result is a plain CoderProvidersFile (no nulls),
+ * which the route re-validates against CoderProvidersFileSchema before save.
+ */
+export function mergeProviderConfigPatch(
+  current: CoderProvidersFile,
+  patch: ProviderConfigPatch,
+): CoderProvidersFile {
+  const providers: Record<string, ProviderOverride> = { ...current.providers };
+  for (const [id, override] of Object.entries(patch.providers)) {
+    if (override === null) {
+      delete providers[id];
+    } else {
+      providers[id] = override;
+    }
+  }
+  return { providers };
+}
+
 /** Read + parse + validate. Falls back to built-ins-only on any failure; never throws. */
 export function load(path: string): CoderProvidersFile {
   let raw: string;

apps/coder/src/services/provider-diagnostic.ts

@@ -0,0 +1,71 @@
+/**
+ * v2.3 Phase 4 (design.md §8) — per-provider plaintext diagnostic report.
+ *
+ * Read-only by default: reports CACHED state (resolved registry def + the
+ * available_agents row + the warm snapshot-cache entry) plus a `which`-style
+ * PATH check for the launch binary. It does NOT spawn an ACP probe — §8 lists
+ * the live initialize probe as optional, and the route defaults to cached state.
+ *
+ * A template string is the whole formatter (no Paseo diagnostic-utils port).
+ */
+import type { ResolvedProviderDef } from './provider-config-registry.js';
+import type { ProviderSnapshotEntry, ProviderModel } from './provider-types.js';
+import { isCommandAvailable } from './command-availability.js';
+
+/** The subset of an `available_agents` row the diagnostic reads. */
+export interface DiagnosticAgentRow {
+  name: string;
+  install_path: string | null;
+  supports_acp?: boolean;
+  models?: ProviderModel[] | null;
+  last_probed_at?: string | Date | null;
+}
+
+interface DiagnosticOpts {
+  /** Warm snapshot-cache entry (read-only peek) — source of the last probe error. */
+  cachedEntry?: ProviderSnapshotEntry;
+  /** Injectable PATH check (defaults to the real `which`); stubbed in tests. */
+  checkAvailable?: (binary: string) => Promise<boolean>;
+}
+
+/** Resolve the binary the dispatcher would launch (for the PATH check + report). */
+function resolveBinary(resolved: ResolvedProviderDef, agentRow: DiagnosticAgentRow | undefined): string {
+  return resolved.launchCommand?.[0] ?? agentRow?.install_path ?? resolved.id;
+}
+
+export async function getProviderDiagnostic(
+  resolved: ResolvedProviderDef,
+  agentRow: DiagnosticAgentRow | undefined,
+  opts: DiagnosticOpts = {},
+): Promise<string> {
+  const checkAvailable = opts.checkAvailable ?? isCommandAvailable;
+  const installed = agentRow?.install_path != null;
+  const binary = resolveBinary(resolved, agentRow);
+  // boocode is native (no binary to launch) — short-circuit the PATH check.
+  const commandAvailable = resolved.transport === 'native' ? true : await checkAvailable(binary);
+  const lastProbedAt =
+    agentRow?.last_probed_at != null ? new Date(agentRow.last_probed_at).toISOString() : '(never)';
+  const modelCount = agentRow?.models?.length ?? 0;
+  const launchCommand = resolved.launchCommand
+    ? resolved.launchCommand.join(' ')
+    : '(built-in default, resolved at dispatch)';
+  const lastError = opts.cachedEntry?.error ?? '(none recorded)';
+
+  return [
+    `provider: ${resolved.id}`,
+    `label: ${resolved.configLabel ?? resolved.label}`,
+    `transport: ${resolved.transport}`,
+    `enabled: ${resolved.enabled}`,
+    `builtin: ${resolved.isBuiltin}`,
+    `customAcp: ${resolved.isCustomAcp}`,
+    `installed: ${installed}`,
+    `install_path: ${agentRow?.install_path ?? '(none)'}`,
+    `binary: ${binary}`,
+    `command_available: ${commandAvailable}`,
+    `launch_command: ${launchCommand}`,
+    `supports_acp: ${agentRow?.supports_acp ?? '(unknown)'}`,
+    `last_probed_at: ${lastProbedAt}`,
+    `models_in_db: ${modelCount}`,
+    `last_probe_error: ${lastError}`,
+  ].join('\n');
+}

apps/coder/src/services/provider-registry.ts

@@ -41,9 +41,18 @@ export const PROVIDERS: ProviderDef[] = [
     label: 'Claude Code',
     transport: 'pty',
     modelSource: 'static',
+    // Passed verbatim to `claude --model <id>` (PTY dispatch). The CLI accepts a
+    // latest-alias ('opus'/'sonnet'/'haiku') or a pinned full name
+    // ('claude-opus-4-8'). Aliases never go stale; pinned IDs let you select an
+    // exact version. Extend/replace per-install via data/coder-providers.json
+    // (models / additionalModels) without a code change.
     staticModels: [
-      { id: 'claude-opus-4-20250514', label: 'Opus 4' },
+      { id: 'opus', label: 'Opus (latest)' },
-      { id: 'claude-sonnet-4-20250514', label: 'Sonnet 4' },
+      { id: 'claude-opus-4-8', label: 'Opus 4.8' },
+      { id: 'sonnet', label: 'Sonnet (latest)' },
+      { id: 'claude-sonnet-4-6', label: 'Sonnet 4.6' },
+      { id: 'haiku', label: 'Haiku (latest)' },
+      { id: 'claude-haiku-4-5-20251001', label: 'Haiku 4.5' },
     ],
   },
   {

apps/coder/src/services/provider-snapshot.ts

@@ -11,23 +11,25 @@ import {
   PROVIDER_MANIFEST,
 } from './provider-manifest.js';
 import { probeAcpProvider } from './acp-probe.js';
-import type { ProviderModel, ProviderSnapshotEntry } from './provider-types.js';
+import type { ProviderModel, ProviderSnapshotEntry, AgentCommand } from './provider-types.js';
 import { getManifestCommands, mergeCommands } from './provider-commands.js';
 import { readQwenSettingsModels } from './qwen-settings.js';
 import { getResolvedRegistry, type ResolvedProviderDef } from './provider-config-registry.js';
 import { isCommandAvailable } from './command-availability.js';
+import { discoverClaudeCommands } from './claude-command-discovery.js';
 
 interface AgentRow {
   name: string;
   install_path: string | null;
   supports_acp: boolean;
   models: ProviderModel[] | null;
+  commands: AgentCommand[] | null;
   label: string | null;
   transport: string | null;
   last_probed_at: string | Date | null;
 }
 
-async function fetchLlamaSwapModels(config: Config): Promise<ProviderModel[]> {
+export async function fetchLlamaSwapModels(config: Config): Promise<ProviderModel[]> {
   try {
     const res = await fetch(`${config.LLAMA_SWAP_URL}/v1/models`);
     if (!res.ok) return [];
@@ -82,9 +84,22 @@ async function buildProviderEntry(
   const fallbackModes = getManifestModes(name);
   const defaultModeId = getManifestDefaultModeId(name);
   const manifestCommands = getManifestCommands(name);
+  // Manifest + persisted live ACP commands (captured on a prior cold probe), so
+  // the agent's discovered commands show even when the tier-2 probe is skipped.
+  const dbCommands = mergeCommands(manifestCommands, agentRow?.commands ?? []);
   const label = agentRow?.label ?? resolved.configLabel ?? resolved.label;
   const descr = resolved.configDescription ? { description: resolved.configDescription } : {};
 
+  // v2.3: config `models` REPLACES the discovered/static list; `additionalModels`
+  // MERGES on top. Applied to every ready/installed model list below.
+  const withConfigModels = (m: ProviderModel[]): ProviderModel[] => {
+    let out = resolved.configModels && resolved.configModels.length > 0 ? resolved.configModels : m;
+    if (resolved.configAdditionalModels && resolved.configAdditionalModels.length > 0) {
+      out = mergeModels(out, resolved.configAdditionalModels);
+    }
+    return out;
+  };
+
   // ACP built-ins fall back to PTY transport when the installed binary lacks ACP.
   let transport = resolved.transport;
   if (agentRow && resolved.transport === 'acp' && !agentRow.supports_acp) {
@@ -104,7 +119,7 @@ async function buildProviderEntry(
   if (isNative) {
     return {
       name, label: resolved.label, transport, status: 'ready',
-      enabled: true, installed: true, models: llamaModels, modes: [],
+      enabled: true, installed: true, models: withConfigModels(llamaModels), modes: [],
       defaultModeId: null, commands: manifestCommands,
     };
   }
@@ -135,10 +150,12 @@ async function buildProviderEntry(
 
   // claude: static models + thinking options, no ACP probe (unchanged from v2.2).
   if (name === 'claude') {
+    // claude is PTY (no ACP discovery) — read its enabled commands + plugin
+    // skills from disk live (the snapshot cache rate-limits the fs reads).
     return {
       name, label, transport, status: 'ready', enabled: true, installed: true,
-      models: attachClaudeThinking(models), modes: fallbackModes, defaultModeId,
+      models: attachClaudeThinking(withConfigModels(models)), modes: fallbackModes, defaultModeId,
-      commands: manifestCommands,
+      commands: mergeCommands(manifestCommands, discoverClaudeCommands()),
     };
   }
 
@@ -165,7 +182,7 @@ async function buildProviderEntry(
       }
       return {
         name, label, transport, status: 'ready', enabled: true, installed: true,
-        models: skipModels, modes: fallbackModes, defaultModeId, commands: manifestCommands,
+        models: withConfigModels(skipModels), modes: fallbackModes, defaultModeId, commands: dbCommands,
       };
     }
 
@@ -188,7 +205,7 @@ async function buildProviderEntry(
       name, label, transport,
       status: probe.ok ? 'ready' : 'error',
       enabled: true, installed: true,
-      models: probeModels,
+      models: withConfigModels(probeModels),
       modes: probe.modes.length > 0 ? probe.modes : fallbackModes,
       defaultModeId: probe.defaultModeId ?? defaultModeId,
       commands: mergeCommands(manifestCommands, probe.commands),
@@ -203,27 +220,10 @@ async function buildProviderEntry(
   }
   return {
     name, label, transport, status: 'ready', enabled: true, installed: true,
-    models, modes: fallbackModes, defaultModeId, commands: manifestCommands,
+    models: withConfigModels(models), modes: fallbackModes, defaultModeId, commands: dbCommands,
   };
 }
 
-/** Synchronous placeholder entries for a cache-miss while the build runs (§4.4). */
-function loadingEntries(): ProviderSnapshotEntry[] {
-  return [...getResolvedRegistry().values()].map((r) => ({
-    name: r.id,
-    label: r.configLabel ?? r.label,
-    ...(r.configDescription ? { description: r.configDescription } : {}),
-    transport: r.transport,
-    status: r.enabled ? ('loading' as const) : ('unavailable' as const),
-    enabled: r.enabled,
-    installed: false,
-    models: [],
-    modes: getManifestModes(r.id),
-    defaultModeId: getManifestDefaultModeId(r.id),
-    commands: getManifestCommands(r.id),
-  }));
-}
-
 const snapshotCache = new Map<string, { at: number; entries: ProviderSnapshotEntry[] }>();
 const snapshotInflight = new Map<string, Promise<ProviderSnapshotEntry[]>>();
 const CACHE_TTL_MS = 5 * 60_000;
@@ -249,7 +249,7 @@ export async function getProviderSnapshot(
   const build = async (): Promise<ProviderSnapshotEntry[]> => {
     const llamaModels = await fetchLlamaSwapModels(config);
     const agents = await sql<AgentRow[]>`
-      SELECT name, install_path, supports_acp, models, label, transport, last_probed_at FROM available_agents
+      SELECT name, install_path, supports_acp, models, commands, label, transport, last_probed_at FROM available_agents
     `;
     const agentMap = new Map(agents.map((a) => [a.name, a]));
     const ttlMs = config.PROVIDER_PROBE_TTL_MS;
@@ -269,14 +269,13 @@ export async function getProviderSnapshot(
   });
   snapshotInflight.set(cacheKey, promise);
 
-  // force → await the full build (cold probes included). Non-force cache miss →
+  // Await the build (force or cache-miss) and return terminal entries. The sync
-  // return loading entries synchronously; the build settles in the background
+  // `loading` return (design §4.4) is DEFERRED until Phase 5 ships the client
-  // and the next call returns it via cache / inflight (§4.4; client polls).
+  // poll that resolves it: without that poll, a single fetch lands on
-  if (force) return promise;
+  // installed:false `loading` entries, which AgentComposerBar filters out
-  promise.catch(() => {
+  // (`e.installed && ...`) → empty picker. Builds stay fast via the tier-2 skip
-    /* settled errors surface on the next call that awaits inflight/rebuilds */
+  // once available_agents.models is warm.
-  });
+  return promise;
-  return loadingEntries();
 }
 
 export function clearProviderSnapshotCache(): void {
@@ -284,6 +283,16 @@ export function clearProviderSnapshotCache(): void {
   snapshotInflight.clear();
 }
 
+/**
+ * Read-only peek into the warm snapshot cache for one provider (no build, no
+ * probe). Used by the diagnostic route to report the last computed probe error
+ * without spawning anything. Returns undefined on a cold cache / unknown name.
+ */
+export function peekSnapshotEntry(name: string, cwd?: string): ProviderSnapshotEntry | undefined {
+  const resolvedCwd = cwd?.trim() || homedir();
+  return snapshotCache.get(resolvedCwd)?.entries.find((e) => e.name === name);
+}
+
 /** Persist probed model lists back to available_agents for fast legacy reads. */
 export async function persistProbedModels(
   sql: Sql,
@@ -292,16 +301,34 @@ export async function persistProbedModels(
 ): Promise<void> {
   let count = 0;
   for (const entry of entries) {
-    if (entry.name === 'boocode' || entry.models.length === 0) continue;
+    if (entry.name === 'boocode') continue;
-    const flatModels = entry.models.map(({ id, label }) => ({ id, label }));
+    let persisted = false;
-    await sql`
+    if (entry.models.length > 0) {
-      UPDATE available_agents
+      const flatModels = entry.models.map(({ id, label }) => ({ id, label }));
-      SET models = ${sql.json(flatModels as never)}, last_probed_at = clock_timestamp()
+      await sql`
-      WHERE name = ${entry.name}
+        UPDATE available_agents
-    `;
+        SET models = ${sql.json(flatModels as never)}, last_probed_at = clock_timestamp()
-    count++;
+        WHERE name = ${entry.name}
+      `;
+      persisted = true;
+    }
+    // Persist captured ACP commands so they survive the tier-2 probe skip and
+    // show without a dispatch. Only when non-empty — never clobber a prior set.
+    if (entry.commands.length > 0) {
+      const flatCommands = entry.commands.map((c) => ({
+        name: c.name,
+        ...(c.description ? { description: c.description } : {}),
+      }));
+      await sql`
+        UPDATE available_agents
+        SET commands = ${sql.json(flatCommands as never)}, last_probed_at = clock_timestamp()
+        WHERE name = ${entry.name}
+      `;
+      persisted = true;
+    }
+    if (persisted) count++;
   }
   if (count > 0) {
-    log.info({ count }, 'provider-snapshot: persisted models to available_agents');
+    log.info({ count }, 'provider-snapshot: persisted models/commands to available_agents');
   }
 }

apps/coder/src/services/provider-types.ts

@@ -30,6 +30,9 @@ export type ProviderSnapshotStatus = 'loading' | 'ready' | 'unavailable' | 'erro
 export interface AgentCommand {
   name: string;
   description?: string;
+  // v2.5.11: 'skill' (plugin skill) vs 'command' (native/CLI slash command).
+  // Drives the icon split in the coder slash menu. Undefined → command.
+  kind?: 'command' | 'skill';
 }
 
 // KEEP IN SYNC with apps/web/src/api/types.ts ProviderSnapshotEntry — parity is

apps/coder/src/services/worktrees.ts

@@ -6,6 +6,7 @@
  * After the agent completes, we diff the worktree against HEAD and
  * queue the diff into pending_changes.
  */
+import type { Sql } from '../db.js';
 import { hostExec } from './host-exec.js';
 
 const WORKTREE_BASE = '/tmp/booworktrees';
@@ -45,7 +46,7 @@ export async function createWorktree(
 export async function diffWorktree(
   worktreePath: string,
   projectPath: string,
-  opts?: { signal?: AbortSignal },
+  opts?: { signal?: AbortSignal; baseRef?: string },
 ): Promise<string> {
   // First, commit any uncommitted changes in the worktree so we can diff branches
   // Stage all changes
@@ -74,9 +75,13 @@ export async function diffWorktree(
     { signal: opts?.signal, timeoutMs: 15_000 },
   );
 
-  // Diff the worktree branch against the parent commit (HEAD of main tree)
+  // Diff the worktree branch against the baseline. Per-task callers default to the
+  // main tree's current HEAD; the session-worktree (opencode) path passes the
+  // captured base_commit so the accumulated diff is stable across turns even if
+  // project HEAD advances.
+  const baseRef = opts?.baseRef ?? 'HEAD';
   const diffResult = await hostExec(
-    `git -C ${shellEscape(projectPath)} diff HEAD...$(git -C ${shellEscape(worktreePath)} rev-parse HEAD)`,
+    `git -C ${shellEscape(projectPath)} diff ${shellEscape(baseRef)}...$(git -C ${shellEscape(worktreePath)} rev-parse HEAD)`,
     { signal: opts?.signal, timeoutMs: 60_000 },
   );
 
@@ -111,6 +116,72 @@ export async function cleanupWorktree(
   ).catch(() => {});
 }
 
+// ─── v2.6: session-keyed persistent worktree ────────────────────────────────
+
+export interface SessionWorktree {
+  worktreePath: string;
+  baseCommit: string | null;
+}
+
+/**
+ * v2.6: create-or-reuse ONE worktree per BooCode session (shared across all
+ * agents/turns in the session), recorded in `session_worktrees`. Unlike the
+ * per-task `createWorktree`, this persists — it is NOT torn down per turn
+ * (cleanup is Phase 3). Captures the project's current HEAD as `base_commit`
+ * so the accumulating diff has a stable baseline across turns.
+ *
+ * Distinct path namespace (`session-<id>` branch, `/sess-<id>` dir) so it never
+ * collides with the per-task worktrees that arena/new_task/MCP still use.
+ */
+export async function ensureSessionWorktree(
+  sql: Sql,
+  projectPath: string,
+  sessionId: string,
+  opts?: { signal?: AbortSignal },
+): Promise<SessionWorktree> {
+  const [existing] = await sql<{ worktree_path: string; base_commit: string | null }[]>`
+    SELECT worktree_path, base_commit FROM session_worktrees WHERE session_id = ${sessionId}
+  `;
+  if (existing) {
+    return { worktreePath: existing.worktree_path, baseCommit: existing.base_commit };
+  }
+
+  const worktreePath = `${WORKTREE_BASE}/sess-${sessionId}`;
+  const branchName = `session-${sessionId}`;
+
+  await hostExec(`mkdir -p ${WORKTREE_BASE}`, { signal: opts?.signal });
+
+  // Capture the baseline commit BEFORE branching, so the diff is stable even if
+  // project HEAD later advances.
+  const headResult = await hostExec(
+    `git -C ${shellEscape(projectPath)} rev-parse HEAD`,
+    { signal: opts?.signal, timeoutMs: 10_000 },
+  );
+  const baseCommit = headResult.exitCode === 0 ? headResult.stdout.trim() || null : null;
+
+  const result = await hostExec(
+    `git -C ${shellEscape(projectPath)} worktree add ${shellEscape(worktreePath)} -b ${shellEscape(branchName)} HEAD`,
+    { signal: opts?.signal, timeoutMs: 30_000 },
+  );
+  if (result.exitCode !== 0) {
+    throw new Error(`Failed to create session worktree: ${result.stderr.trim() || result.stdout.trim()}`);
+  }
+
+  // Persist. ON CONFLICT keeps the first writer's row if two turns race the create.
+  await sql`
+    INSERT INTO session_worktrees (session_id, worktree_path, base_commit)
+    VALUES (${sessionId}, ${worktreePath}, ${baseCommit})
+    ON CONFLICT (session_id) DO NOTHING
+  `;
+  const [row] = await sql<{ worktree_path: string; base_commit: string | null }[]>`
+    SELECT worktree_path, base_commit FROM session_worktrees WHERE session_id = ${sessionId}
+  `;
+  return {
+    worktreePath: row?.worktree_path ?? worktreePath,
+    baseCommit: row?.base_commit ?? baseCommit,
+  };
+}
+
 /** Minimal shell escape for paths (single-quote wrapping). */
 function shellEscape(s: string): string {
   // Replace single quotes with escaped version, wrap in single quotes

apps/server/src/services/auto_name.ts

@@ -37,9 +37,11 @@ export async function maybeAutoNameChat(
   if ((counts[0]?.n ?? 0) < 1) return;
 
   const chatRows = await ctx.sql<
-    { id: string; name: string | null; session_id: string }[]
+    { id: string; name: string | null; session_id: string; model: string | null }[]
   >`
-    SELECT id, name, session_id FROM chats WHERE id = ${chatId}
+    SELECT c.id, c.name, c.session_id, s.model
+    FROM chats c JOIN sessions s ON s.id = c.session_id
+    WHERE c.id = ${chatId}
   `;
   const chat = chatRows[0];
   if (!chat) return;
@@ -67,6 +69,7 @@ export async function maybeAutoNameChat(
     user: namingInput,
     maxTokens: 30,
     temperature: 0.3,
+    fallbackModel: chat.model ?? undefined,
   });
   const name = cleanTitle(raw);
   if (!name) {

apps/web/src/api/client.ts

@@ -14,6 +14,8 @@ import type {
   AskUserAnswer,
   ToolCostStat,
   ProviderSnapshotEntry,
+  CoderProvidersFile,
+  ProviderConfigPatch,
   CoderSendMessageBody,
   CoderSendMessageResponse,
   CoderMessageWire,
@@ -310,8 +312,23 @@ export const api = {
       const qs = cwd ? `?cwd=${encodeURIComponent(cwd)}` : '';
       return request<ProviderSnapshotEntry[]>(`/api/coder/providers/snapshot${qs}`);
     },
-    refreshProviders: () =>
+    // v2.3 Phase 4: optional subset narrows the reported `refreshed` count.
-      request<{ refreshed: number }>('/api/coder/providers/refresh', { method: 'POST' }),
+    refreshProviders: (providers?: string[]) =>
+      request<{ refreshed: number }>('/api/coder/providers/refresh', {
+        method: 'POST',
+        ...(providers && providers.length > 0 ? { body: JSON.stringify({ providers }) } : {}),
+      }),
+    // v2.3 Phase 4: read/patch the provider config file. PATCH returns the new
+    // config; a `null` value in the patch deletes that id's override.
+    getProvidersConfig: () => request<CoderProvidersFile>('/api/coder/providers/config'),
+    patchProvidersConfig: (patch: ProviderConfigPatch) =>
+      request<{ ok: true } & CoderProvidersFile>('/api/coder/providers/config', {
+        method: 'PATCH',
+        body: JSON.stringify(patch),
+      }),
+    // v2.3 Phase 4: per-provider diagnostic — JSON { diagnostic: string } (§6.4).
+    getProviderDiagnostic: (id: string) =>
+      request<{ diagnostic: string }>(`/api/coder/providers/${encodeURIComponent(id)}/diagnostic`),
     sendMessage: (sessionId: string, body: CoderSendMessageBody) =>
       request<CoderSendMessageResponse>(`/api/coder/sessions/${sessionId}/messages`, {
         method: 'POST',
@@ -332,18 +349,32 @@ export const api = {
       request<CoderMessageWire[]>(
         `/api/coder/sessions/${sessionId}/messages${chatId ? `?chat_id=${encodeURIComponent(chatId)}` : ''}`,
       ),
-    skillInvoke: (sessionId: string, paneId: string, skillName: string, userMessage: string | null) =>
+    skillInvoke: (
+      sessionId: string,
+      paneId: string,
+      skillName: string,
+      userMessage: string | null,
+      // v2.5.9: when the active provider is external, the skill runs under that
+      // agent (body injected into a dispatched task) → response carries task_id.
+      config?: { provider?: string; model?: string; mode_id?: string; thinking_option_id?: string },
+    ) =>
       request<{
         user_message_id: string;
-        assistant_message_id: string;
+        assistant_message_id?: string;
-        synth_assistant_id: string;
+        synth_assistant_id?: string;
-        tool_message_id: string;
+        tool_message_id?: string;
+        task_id?: string;
+        dispatched?: boolean;
       }>(`/api/coder/sessions/${sessionId}/skill_invoke`, {
         method: 'POST',
         body: JSON.stringify({
           pane_id: paneId,
           skill_name: skillName,
           user_message: userMessage,
+          ...(config?.provider ? { provider: config.provider } : {}),
+          ...(config?.model ? { model: config.model } : {}),
+          ...(config?.mode_id ? { mode_id: config.mode_id } : {}),
+          ...(config?.thinking_option_id ? { thinking_option_id: config.thinking_option_id } : {}),
         }),
       }),
     // Queue a new-file create from the RightRail browser → BooCoder

apps/web/src/api/types.ts

@@ -253,6 +253,31 @@ export interface ProviderSnapshotEntry {
   fetchedAt?: string;
 }
 
+// v2.3 Phase 4: provider config file wire types. Mirror of the Zod-inferred
+// ProviderOverride / CoderProvidersFile in apps/coder/src/services/provider-config.ts
+// (web can't cross-import the coder package — TS6307 on the composite project).
+export interface ProviderOverride {
+  extends?: 'acp';
+  label?: string;
+  description?: string;
+  command?: string[];
+  env?: Record<string, string>;
+  enabled?: boolean;
+  order?: number;
+  models?: Array<{ id: string; label: string }>;
+  additionalModels?: Array<{ id: string; label: string }>;
+}
+
+export interface CoderProvidersFile {
+  providers: Record<string, ProviderOverride>;
+}
+
+// PATCH body: a partial providers map. A `null` value deletes that id's
+// override (revert to built-in default); an object replaces it wholesale.
+export interface ProviderConfigPatch {
+  providers: Record<string, ProviderOverride | null>;
+}
+
 export interface AgentSessionConfig {
   provider: string;
   model: string;
@@ -273,6 +298,9 @@ export interface PermissionPrompt {
 export interface AgentCommand {
   name: string;
   description?: string;
+  // v2.5.11: 'skill' (plugin skill) vs 'command' (native/CLI slash command).
+  // Drives the icon split in the coder slash menu. Undefined → command.
+  kind?: 'command' | 'skill';
 }
 
 export interface CoderSendMessageBody {

apps/web/src/components/AgentComposerBar.tsx

@@ -1,5 +1,5 @@
 import { useEffect, useMemo, useRef, useState } from 'react';
-import { Check, ChevronDown, RefreshCw, Shield, Brain, Bird, Bot, Dog, Terminal as TermIcon } from 'lucide-react';
+import { Check, ChevronDown, RefreshCw, Loader2, Shield, Brain, Bird, Bot, Dog, Terminal as TermIcon } from 'lucide-react';
 import { ClaudeIcon, OpenCodeIcon } from '@/components/icons/ProviderIcons';
 import { api } from '@/api/client';
 import type { AgentSessionConfig, ProviderSnapshotEntry, AgentCommand } from '@/api/types';
@@ -92,9 +92,11 @@ interface PickerProps {
   options: Array<{ id: string; label: string }>;
   onPick: (id: string) => void;
   icon?: React.ReactNode;
+  /** Mobile: render icon + chevron only (no value label) to save row width. */
+  iconOnly?: boolean;
 }
 
-function CompactPicker({ label, value, disabled, options, onPick, icon }: PickerProps) {
+function CompactPicker({ label, value, disabled, options, onPick, icon, iconOnly }: PickerProps) {
   const { isMobile } = useViewport();
   const [open, setOpen] = useState(false);
   const currentLabel = options.find((o) => o.id === value)?.label ?? (value || label);
@@ -129,7 +131,7 @@ function CompactPicker({ label, value, disabled, options, onPick, icon }: Picker
           className="inline-flex items-center gap-1 min-h-[44px] px-1.5 rounded text-xs text-muted-foreground hover:text-foreground disabled:opacity-40"
         >
           {icon}
-          <span className="truncate max-w-[120px]">{currentLabel}</span>
+          {!iconOnly && <span className="truncate max-w-[120px]">{currentLabel}</span>}
           <ChevronDown className="size-3 opacity-70 shrink-0" />
         </button>
         <BottomSheet open={open} onClose={() => setOpen(false)} title={label}>
@@ -174,8 +176,12 @@ interface Props {
 
 export function AgentComposerBar({ projectPath, value, onChange, onProviderCommandsChange, connected }: Props) {
   const allEntries = useProviderSnapshot(projectPath);
+  // 5.5 — the composer picker only offers ENABLED providers that are ready (or
+  // still loading). Disabled (enabled:false) and unavailable/error providers are
+  // hidden here and managed in Settings → Providers. Native boocode is always
+  // enabled+ready, so it always appears.
   const entries = useMemo(
-    () => allEntries?.filter((e) => e.installed && e.status !== 'error') ?? null,
+    () => allEntries?.filter((e) => e.enabled && (e.status === 'ready' || e.status === 'loading')) ?? null,
     [allEntries],
   );
   const [refreshing, setRefreshing] = useState(false);
@@ -198,6 +204,35 @@ export function AgentComposerBar({ projectPath, value, onChange, onProviderComma
     onChange(resolveConfig(entry, prefs));
   }, [entries, onChange, value.provider]);
 
+  // If the active provider is disabled in the settings drawer it drops out of
+  // `entries` (the 5.5 filter) — fall back to boocode so the composer never
+  // strands on an unselectable provider with empty model/mode pickers.
+  useEffect(() => {
+    if (!entries?.length) return;
+    if (entries.some((e) => e.name === value.provider)) return;
+    const fallback = entries.find((e) => e.name === 'boocode') ?? entries[0];
+    if (!fallback) return;
+    onChange(resolveConfig(fallback, loadPrefs()));
+  }, [entries, value.provider, onChange]);
+
+  // 5.6 — loading poll: while any entry is loading (Phase 2's sync cache-miss
+  // return), refetch until terminal. Capped; no provider_snapshot_updated WS
+  // frame (deferred Tier-2). Dormant today since the snapshot awaits the build.
+  const pollsRef = useRef(0);
+  useEffect(() => {
+    const anyLoading = allEntries?.some((e) => e.status === 'loading') ?? false;
+    if (!anyLoading) {
+      pollsRef.current = 0;
+      return;
+    }
+    if (pollsRef.current >= 10) return;
+    const t = setTimeout(() => {
+      pollsRef.current += 1;
+      void refreshProviderSnapshot(projectPath);
+    }, 2000);
+    return () => clearTimeout(t);
+  }, [allEntries, projectPath]);
+
   const currentEntry = useMemo(
     () => entries?.find((e) => e.name === value.provider),
     [entries, value.provider],
@@ -281,7 +316,11 @@ export function AgentComposerBar({ projectPath, value, onChange, onProviderComma
         value={value.provider}
         options={providerOptions}
         onPick={pickProvider}
-        icon={providerIcon(value.provider)}
+        icon={
+          currentEntry?.status === 'loading'
+            ? <Loader2 size={13} className="shrink-0 animate-spin" />
+            : providerIcon(value.provider)
+        }
       />
       <CompactPicker
         label="Mode"
@@ -290,6 +329,7 @@ export function AgentComposerBar({ projectPath, value, onChange, onProviderComma
         options={modeOptions}
         onPick={(modeId) => persist({ ...value, modeId })}
         icon={<Shield className="size-3 shrink-0" />}
+        iconOnly
       />
       <CompactPicker
         label="Model"
@@ -308,22 +348,26 @@ export function AgentComposerBar({ projectPath, value, onChange, onProviderComma
           icon={<Brain className="size-3 shrink-0" />}
         />
       )}
-      {connected !== undefined && (
+      {/* Status dot + refresh as one right-aligned unit so the refresh button
-        <span
+          stays on the top line instead of wrapping past the edge-pinned dot. */}
-          className={cn('inline-block w-1.5 h-1.5 rounded-full shrink-0 ml-auto', connected ? 'bg-green-500' : 'bg-red-500')}
+      <div className="ml-auto flex items-center gap-1 shrink-0">
-          title={connected ? 'Connected' : 'Disconnected'}
+        {connected !== undefined && (
-        />
+          <span
-      )}
+            className={cn('inline-block w-1.5 h-1.5 rounded-full shrink-0', connected ? 'bg-green-500' : 'bg-red-500')}
-      <button
+            title={connected ? 'Connected' : 'Disconnected'}
-        type="button"
+          />
-        onClick={() => void handleRefresh()}
+        )}
-        disabled={refreshing}
+        <button
-        className={cn('inline-flex items-center justify-center size-7 max-md:min-h-[44px] max-md:min-w-[44px] rounded text-muted-foreground hover:text-foreground disabled:opacity-40', connected === undefined && 'ml-auto')}
+          type="button"
-        aria-label="Refresh provider list"
+          onClick={() => void handleRefresh()}
-        title="Refresh providers"
+          disabled={refreshing}
-      >
+          className="inline-flex items-center justify-center size-7 max-md:min-h-[44px] max-md:min-w-[44px] rounded text-muted-foreground hover:text-foreground disabled:opacity-40"
-        <RefreshCw className={cn('size-3.5', refreshing && 'animate-spin')} />
+          aria-label="Refresh provider list"
-      </button>
+          title="Refresh providers"
+        >
+          <RefreshCw className={cn('size-3.5', refreshing && 'animate-spin')} />
+        </button>
+      </div>
     </div>
   );
 }

apps/web/src/components/ChatInput.tsx

@@ -24,7 +24,7 @@ import { DropOverlay } from '@/components/DropOverlay';
 import { AgentPicker } from '@/components/AgentPicker';
 import { AgentCommandsHint } from '@/components/AgentCommandsHint';
 import { ContextBar } from '@/components/ContextBar';
-import { SlashCommandPicker } from '@/components/SlashCommandPicker';
+import { SlashCommandPicker, type SlashCommandGroup } from '@/components/SlashCommandPicker';
 import { isSlashCommandToken, parseSlashInput, slashQuery } from '@/lib/slash-command';
 import { api } from '@/api/client';
 import type { Message } from '@/api/types';
@@ -56,6 +56,13 @@ interface Props {
   // empty). Callers wire this to api.chats.skillInvoke. Omitting the prop
   // disables slash-command dispatch (input is sent as literal text).
   onSlashCommand?: (skillName: string, userMessage: string) => void | Promise<void>;
+  // v2.5.9: segmented slash-command DISPLAY source for the picker + hint. When
+  // provided (e.g. CoderPane passing [agent commands, skills]), these labeled
+  // groups are shown instead of the BooChat skills. Invocation routing still
+  // uses the skills lookup — names not in skills (opencode's /help etc.) fall
+  // through and are sent to the agent as literal text. Omitted → BooChat skills
+  // (flat, unchanged — parity).
+  slashGroups?: SlashCommandGroup[];
   // v1.10.4: send-to-chat reverse path. When chatId is provided, this input
   // registers in chatInputsRegistry so the terminal floating menu can list
   // it, and subscribes to sendToChat events scoped to this chatId. Receiving
@@ -71,7 +78,7 @@ interface Props {
   modelContextLimit?: number | null;
 }
 
-export function ChatInput({ disabled, projectId, agentId, onAgentChange, sessionId, webSearchEnabled, onSend, onForceSend, onSlashCommand, chatId, chatLabel, messages, modelContextLimit }: Props) {
+export function ChatInput({ disabled, projectId, agentId, onAgentChange, sessionId, webSearchEnabled, onSend, onForceSend, onSlashCommand, slashGroups, chatId, chatLabel, messages, modelContextLimit }: Props) {
   const { isMobile } = useViewport();
   const [value, setValue] = useState('');
   const [busy, setBusy] = useState(false);
@@ -100,6 +107,15 @@ export function ChatInput({ disabled, projectId, agentId, onAgentChange, session
     for (const s of skills) m.set(s.name, true);
     return m;
   }, [skills]);
+  // Flat display source for the hint (and the picker's no-groups fallback):
+  // caller-provided groups flattened, else the BooChat skills.
+  const slashItems = useMemo(
+    () =>
+      slashGroups
+        ? slashGroups.flatMap((g) => g.items)
+        : skills.map((s) => ({ name: s.name, description: s.description })),
+    [slashGroups, skills],
+  );
   const [fileIndex, setFileIndex] = useState<string[] | null>(null);
   const textareaRef = useRef<HTMLTextAreaElement | null>(null);
 
@@ -561,8 +577,8 @@ export function ChatInput({ disabled, projectId, agentId, onAgentChange, session
           ))}
         </div>
       )}
-      {skills.length > 0 && (
+      {slashItems.length > 0 && (
-        <AgentCommandsHint commands={skills.map((s) => ({ name: s.name, description: s.description }))} />
+        <AgentCommandsHint commands={slashItems} />
       )}
       {/* Batch 9 toolbar — agent picker + quick-toggle menu. v1.11.5.1
           inlines ContextBar in the same row so the bar lives next to the
@@ -661,11 +677,12 @@ export function ChatInput({ disabled, projectId, agentId, onAgentChange, session
       {slashState && (
         <SlashCommandPicker
           query={slashState.query}
-          items={skills}
+          items={slashItems}
+          groups={slashGroups}
           inputRef={textareaRef}
           onSelect={handleSlashSelect}
           onClose={() => setSlashState(null)}
-          emptyLabel="No skills available"
+          emptyLabel={slashGroups ? 'No commands available' : 'No skills available'}
         />
       )}
     </div>

apps/web/src/components/ChatTabBar.tsx

@@ -1,5 +1,5 @@
 import { useState } from 'react';
-import { Code, History, MessageSquare, Plus, Terminal, X } from 'lucide-react';
+import { Code, Columns2, History, MessageSquare, Plus, RotateCcw, Terminal, X } from 'lucide-react';
 import type { Chat, WorkspacePane } from '@/api/types';
 import { StatusDot } from '@/components/StatusDot';
 import {
@@ -26,7 +26,9 @@ interface Props {
   onCloseOthers: (chatId: string) => void;
   onCloseToRight: (chatId: string) => void;
   onCloseAll: () => void;
-  onAddPane: (kind: 'chat' | 'terminal' | 'coder') => void;
+  onNewTab: () => void;
+  onSplitPane: (kind: 'chat' | 'terminal' | 'coder') => void;
+  onReopenPane?: () => void;
   onShowHistory: () => void;
   onRename: (chatId: string, name: string) => Promise<void>;
   onRemovePane?: () => void;
@@ -40,7 +42,9 @@ export function ChatTabBar({
   onCloseOthers,
   onCloseToRight,
   onCloseAll,
-  onAddPane,
+  onNewTab,
+  onSplitPane,
+  onReopenPane,
   onShowHistory,
   onRename,
   onRemovePane,
@@ -131,7 +135,7 @@ export function ChatTabBar({
               </div>
             </ContextMenuTrigger>
             <ContextMenuContent>
-              <ContextMenuItem onSelect={() => onAddPane('chat')}>
+              <ContextMenuItem onSelect={onNewTab}>
                 New chat
               </ContextMenuItem>
               <ContextMenuSeparator />
@@ -170,29 +174,49 @@ export function ChatTabBar({
       )}
 
       <div className="flex items-center ml-auto gap-0.5 px-1 shrink-0">
+        <button
+          type="button"
+          onClick={onNewTab}
+          className="inline-flex items-center justify-center p-1 rounded text-muted-foreground hover:bg-muted hover:text-foreground max-md:min-h-[44px] max-md:min-w-[44px]"
+          aria-label="New tab"
+          title="New tab"
+        >
+          <Plus size={12} />
+        </button>
         <DropdownMenu>
           <DropdownMenuTrigger asChild>
             <button
               type="button"
               className="inline-flex items-center justify-center p-1 rounded text-muted-foreground hover:bg-muted hover:text-foreground max-md:min-h-[44px] max-md:min-w-[44px]"
-              aria-label="New pane"
+              aria-label="Split pane"
-              title="New pane"
+              title="Split pane"
             >
-              <Plus size={12} />
+              <Columns2 size={12} />
             </button>
           </DropdownMenuTrigger>
           <DropdownMenuContent align="end" className="w-fit">
-            <DropdownMenuItem onSelect={() => onAddPane('chat')}>
+            <DropdownMenuItem onSelect={() => onSplitPane('chat')}>
               <MessageSquare size={14} /> New BooChat
             </DropdownMenuItem>
-            <DropdownMenuItem onSelect={() => onAddPane('terminal')}>
+            <DropdownMenuItem onSelect={() => onSplitPane('terminal')}>
               <Terminal size={14} /> New BooTerm
             </DropdownMenuItem>
-            <DropdownMenuItem onSelect={() => onAddPane('coder')}>
+            <DropdownMenuItem onSelect={() => onSplitPane('coder')}>
               <Code size={14} /> New BooCode
             </DropdownMenuItem>
           </DropdownMenuContent>
         </DropdownMenu>
+        {onReopenPane && (
+          <button
+            type="button"
+            onClick={onReopenPane}
+            className="inline-flex items-center justify-center p-1 rounded text-muted-foreground hover:bg-muted hover:text-foreground max-md:min-h-[44px] max-md:min-w-[44px]"
+            aria-label="Reopen closed pane"
+            title="Reopen closed pane"
+          >
+            <RotateCcw size={12} />
+          </button>
+        )}
         <button
           type="button"
           onClick={onShowHistory}

apps/web/src/components/SessionLandingPage.tsx

@@ -1,6 +1,5 @@
 import { useCallback, useState } from 'react';
 import { toast } from 'sonner';
-import { api } from '@/api/client';
 import { ChatInput } from '@/components/ChatInput';
 
 interface Props {
@@ -9,6 +8,10 @@ interface Props {
   agentId?: string | null;
   onAgentChange?: (agentId: string | null) => void | Promise<void>;
   onSend: (content: string) => void;
+  // Slash-command (skill) send from the landing page. The parent creates the
+  // chat, assigns it to the pane (so it transitions to ChatPane), and invokes
+  // the skill — same transition the text send uses. See useSessionChats.
+  onSkillInvoke: (skillName: string, userMessage: string | null) => void;
   createChat: () => Promise<{ id: string }>;
 }
 
@@ -18,6 +21,7 @@ export function SessionLandingPage({
   agentId,
   onAgentChange,
   onSend,
+  onSkillInvoke,
   createChat,
 }: Props) {
   const [chatId, setChatId] = useState<string | null>(null);
@@ -45,14 +49,13 @@ export function SessionLandingPage({
     }
   }, [ensureChat, onSend]);
 
-  const handleSlashCommand = useCallback(async (skillName: string, userMessage: string) => {
+  // Route to the parent, which creates the chat, assigns it to the pane (so the
-    try {
+  // pane transitions to ChatPane and subscribes to the stream), then invokes the
-      const cid = await ensureChat();
+  // skill — mirroring the text-send transition. Doing the skill invoke locally
-      await api.chats.skillInvoke(cid, skillName, userMessage.length > 0 ? userMessage : null);
+  // (without the pane assignment) left the landing pane stuck/blank.
-    } catch (err) {
+  const handleSlashCommand = useCallback((skillName: string, userMessage: string) => {
-      toast.error(err instanceof Error ? err.message : `/${skillName} failed`);
+    onSkillInvoke(skillName, userMessage.length > 0 ? userMessage : null);
-    }
+  }, [onSkillInvoke]);
-  }, [ensureChat]);
 
   return (
     <div className="flex flex-col h-full min-h-0">

apps/web/src/components/SlashCommandPicker.tsx

@@ -1,5 +1,5 @@
 import { useEffect, useMemo, useRef, useState } from 'react';
-import type { CSSProperties, RefObject } from 'react';
+import type { CSSProperties, ReactNode, RefObject } from 'react';
 import { createPortal } from 'react-dom';
 import { cn } from '@/lib/utils';
 
@@ -8,9 +8,19 @@ export interface SlashCommandItem {
   description?: string;
 }
 
+export interface SlashCommandGroup {
+  label: string;
+  items: SlashCommandItem[];
+  icon?: ReactNode;
+}
+
 interface Props {
   query: string;
   items: SlashCommandItem[];
+  // Optional segmented rendering. When provided, items are shown under labeled
+  // group headers (in order). `items` is ignored. BooChat passes only `items`
+  // (flat) so its menu is unchanged — grouping is opt-in.
+  groups?: SlashCommandGroup[];
   inputRef: RefObject<HTMLElement | null>;
   onSelect: (name: string) => void;
   onClose: () => void;
@@ -28,6 +38,7 @@ function filterByPrefix(items: SlashCommandItem[], query: string): SlashCommandI
 export function SlashCommandPicker({
   query,
   items,
+  groups,
   inputRef,
   onSelect,
   onClose,
@@ -35,7 +46,21 @@ export function SlashCommandPicker({
 }: Props) {
   const [highlightIndex, setHighlightIndex] = useState(0);
   const popoverRef = useRef<HTMLDivElement>(null);
-  const filtered = useMemo(() => filterByPrefix(items, query), [items, query]);
+  // When grouped, filter each group and drop empties; otherwise the flat list.
+  const filteredGroups = useMemo(
+    () =>
+      groups
+        ? groups
+            .map((g) => ({ label: g.label, icon: g.icon, items: filterByPrefix(g.items, query) }))
+            .filter((g) => g.items.length > 0)
+        : null,
+    [groups, query],
+  );
+  // Flat list drives keyboard nav + Enter selection across all groups.
+  const filtered = useMemo(
+    () => (filteredGroups ? filteredGroups.flatMap((g) => g.items) : filterByPrefix(items, query)),
+    [filteredGroups, items, query],
+  );
 
   const [rect, setRect] = useState<DOMRect | null>(
     () => inputRef.current?.getBoundingClientRect() ?? null,
@@ -130,6 +155,36 @@ export function SlashCommandPicker({
     // eslint-disable-next-line react-hooks/exhaustive-deps
   }, [rect, vvTick]);
 
+  const renderItem = (item: SlashCommandItem, i: number) => (
+    <div
+      key={`${i}-${item.name}`}
+      role="option"
+      aria-selected={i === highlightIndex}
+      data-highlighted={i === highlightIndex}
+      className={cn(
+        'w-full text-left px-2.5 py-2 cursor-pointer block',
+        i === highlightIndex && 'bg-muted',
+      )}
+      onMouseEnter={() => setHighlightIndex(i)}
+      onClick={() => onSelect(item.name)}
+    >
+      <div className="font-mono text-xs font-bold text-foreground">/{item.name}</div>
+      {item.description && (
+        <div
+          className="text-xs text-muted-foreground overflow-hidden"
+          style={{
+            display: '-webkit-box',
+            WebkitLineClamp: 2,
+            WebkitBoxOrient: 'vertical',
+          }}
+        >
+          {item.description}
+        </div>
+      )}
+    </div>
+  );
+
+  let runningIndex = -1;
   const popover = filtered.length === 0 ? (
     <div
       ref={popoverRef}
@@ -146,34 +201,17 @@ export function SlashCommandPicker({
       className="z-50 bg-popover border border-border rounded-md shadow min-w-[320px] max-w-[420px] max-h-[320px] overflow-y-auto overscroll-contain touch-pan-y"
       style={style}
     >
-      {filtered.map((item, i) => (
+      {filteredGroups
-        <div
+        ? filteredGroups.map((g) => (
-          key={item.name}
+            <div key={g.label}>
-          role="option"
+              <div className="px-2.5 pt-2 pb-1 text-[10px] font-semibold uppercase tracking-wide text-muted-foreground/70 flex items-center gap-1.5">
-          aria-selected={i === highlightIndex}
+                {g.icon}
-          data-highlighted={i === highlightIndex}
+                {g.label}
-          className={cn(
+              </div>
-            'w-full text-left px-2.5 py-2 cursor-pointer block',
+              {g.items.map((item) => renderItem(item, (runningIndex += 1)))}
-            i === highlightIndex && 'bg-muted',
-          )}
-          onMouseEnter={() => setHighlightIndex(i)}
-          onClick={() => onSelect(item.name)}
-        >
-          <div className="font-mono text-xs font-bold text-foreground">/{item.name}</div>
-          {item.description && (
-            <div
-              className="text-xs text-muted-foreground overflow-hidden"
-              style={{
-                display: '-webkit-box',
-                WebkitLineClamp: 2,
-                WebkitBoxOrient: 'vertical',
-              }}
-            >
-              {item.description}
             </div>
-          )}
+          ))
-        </div>
+        : filtered.map((item, i) => renderItem(item, i))}
-      ))}
     </div>
   );
 

apps/web/src/components/Workspace.tsx

@@ -65,6 +65,8 @@ export function Workspace({
     showLandingPage,
     addSplitPane,
     removePane,
+    reopenPane,
+    hasClosedPanes,
     isPaneChatPending,
     handlePaneDragStart,
     handlePaneDragOver,
@@ -82,6 +84,7 @@ export function Workspace({
     deleteChat,
     renameChat,
     handleLandingSend,
+    handleLandingSkill,
   } = chatsHook;
 
   const { isMobile } = useViewport();
@@ -206,10 +209,9 @@ export function Workspace({
                   onCloseOthers={(chatId) => closeOtherTabs(idx, chatId)}
                   onCloseToRight={(chatId) => closeTabsToRight(idx, chatId)}
                   onCloseAll={() => closeAllTabs(idx)}
-                  onAddPane={(kind) => {
+                  onNewTab={() => void createChat(idx)}
-                    if (kind === 'chat') void createChat(idx);
+                  onSplitPane={(kind) => onAddPane(kind)}
-                    else onAddPane(kind);
+                  onReopenPane={hasClosedPanes ? reopenPane : undefined}
-                  }}
                   onShowHistory={() => showLandingPage(idx)}
                   onRename={renameChat}
                   onRemovePane={panes.length > 1 ? () => removePane(idx) : undefined}
@@ -387,6 +389,7 @@ export function Workspace({
                   onAgentChange={onAgentChange}
                   createChat={() => api.chats.create(sessionId)}
                   onSend={(content) => void handleLandingSend(idx, content)}
+                  onSkillInvoke={(skillName, userMessage) => void handleLandingSkill(idx, skillName, userMessage)}
                 />
               )}
             </div>

apps/web/src/components/coder/AddProviderModal.tsx

@@ -0,0 +1,139 @@
+import { useMemo, useState } from 'react';
+import { ExternalLink, Search } from 'lucide-react';
+import { api } from '@/api/client';
+import { Button } from '@/components/ui/button';
+import {
+  Dialog,
+  DialogContent,
+  DialogDescription,
+  DialogFooter,
+  DialogHeader,
+  DialogTitle,
+} from '@/components/ui/dialog';
+import { Input } from '@/components/ui/input';
+import {
+  ACP_PROVIDER_CATALOG,
+  buildAcpProviderConfigPatch,
+  type AcpCatalogEntry,
+} from '@/data/acp-provider-catalog';
+
+interface Props {
+  open: boolean;
+  onOpenChange: (open: boolean) => void;
+  /** Fired after a successful add so the parent can refetch the snapshot. */
+  onAdded: (id: string) => void;
+}
+
+/**
+ * v2.3 Phase 5 (design.md §7.3). Search the curated ACP catalog and register a
+ * provider: PATCH /api/providers/config with its custom-ACP override, then
+ * refresh that one provider. Adding only edits config — it does NOT install the
+ * binary, so the provider shows "Not installed" until the CLI is on PATH.
+ */
+export function AddProviderModal({ open, onOpenChange, onAdded }: Props) {
+  const [query, setQuery] = useState('');
+  const [busyId, setBusyId] = useState<string | null>(null);
+  const [error, setError] = useState<string | null>(null);
+
+  const filtered = useMemo(() => {
+    const q = query.trim().toLowerCase();
+    if (!q) return ACP_PROVIDER_CATALOG;
+    return ACP_PROVIDER_CATALOG.filter(
+      (e) =>
+        e.id.toLowerCase().includes(q) ||
+        e.label.toLowerCase().includes(q) ||
+        e.description.toLowerCase().includes(q),
+    );
+  }, [query]);
+
+  async function add(entry: AcpCatalogEntry): Promise<void> {
+    setBusyId(entry.id);
+    setError(null);
+    try {
+      await api.coder.patchProvidersConfig(buildAcpProviderConfigPatch(entry));
+      await api.coder.refreshProviders([entry.id]);
+      onAdded(entry.id);
+      onOpenChange(false);
+    } catch (err) {
+      // 422 from PATCH (invalid override) surfaces here as ApiError.message.
+      setError(err instanceof Error ? err.message : 'failed to add provider');
+    } finally {
+      setBusyId(null);
+    }
+  }
+
+  return (
+    <Dialog open={open} onOpenChange={onOpenChange}>
+      <DialogContent className="sm:max-w-lg max-h-[85vh] grid-rows-[auto_minmax(0,1fr)_auto]">
+        <DialogHeader>
+          <DialogTitle>Add ACP provider</DialogTitle>
+          <DialogDescription>
+            Registers the provider in your coder config. It is not installed — install the CLI
+            yourself; until it's on PATH it shows as “Not installed”.
+          </DialogDescription>
+        </DialogHeader>
+
+        <div className="flex flex-col min-h-0 gap-3">
+          <div className="relative shrink-0">
+            <Search className="absolute left-2 top-1/2 -translate-y-1/2 size-3.5 text-muted-foreground" />
+            <Input
+              value={query}
+              onChange={(e) => setQuery(e.target.value)}
+              placeholder="Search providers…"
+              className="pl-7"
+            />
+          </div>
+
+          <div className="flex-1 min-h-0 rounded-md border overflow-y-auto overscroll-contain divide-y">
+            {filtered.length === 0 && (
+              <div className="px-3 py-2 text-sm text-muted-foreground">No matching providers.</div>
+            )}
+            {filtered.map((e) => (
+              <div key={e.id} className="px-3 py-2.5 space-y-1.5">
+                <div className="flex items-start justify-between gap-2">
+                  <div className="min-w-0">
+                    <div className="text-sm font-medium">{e.label}</div>
+                    <div className="text-xs text-muted-foreground">{e.description}</div>
+                  </div>
+                  <Button
+                    size="sm"
+                    disabled={busyId !== null}
+                    onClick={() => void add(e)}
+                  >
+                    {busyId === e.id ? 'Adding…' : 'Add'}
+                  </Button>
+                </div>
+                <div className="font-mono text-[11px] text-muted-foreground truncate">
+                  $ {e.command.join(' ')}
+                </div>
+                <div className="flex items-center gap-3">
+                  <a
+                    href={e.installUrl}
+                    target="_blank"
+                    rel="noreferrer"
+                    className="inline-flex items-center gap-1 text-xs text-primary hover:underline"
+                  >
+                    Install {e.label} <ExternalLink className="size-3" />
+                  </a>
+                  {e.installCmd && (
+                    <span className="font-mono text-[11px] text-muted-foreground truncate">
+                      {e.installCmd}
+                    </span>
+                  )}
+                </div>
+              </div>
+            ))}
+          </div>
+
+          {error && <div className="text-sm text-destructive shrink-0">{error}</div>}
+        </div>
+
+        <DialogFooter>
+          <Button variant="outline" onClick={() => onOpenChange(false)} disabled={busyId !== null}>
+            Close
+          </Button>
+        </DialogFooter>
+      </DialogContent>
+    </Dialog>
+  );
+}

apps/web/src/components/coder/ProvidersSettings.tsx

@@ -0,0 +1,218 @@
+import { useEffect, useRef, useState } from 'react';
+import { Loader2, Plus, RefreshCw, Stethoscope } from 'lucide-react';
+import { api } from '@/api/client';
+import type { CoderProvidersFile, ProviderOverride, ProviderSnapshotEntry } from '@/api/types';
+import { useProviderSnapshot, refreshProviderSnapshot } from '@/hooks/useProviderSnapshot';
+import { Button } from '@/components/ui/button';
+import { AddProviderModal } from './AddProviderModal';
+import { cn } from '@/lib/utils';
+
+/** Map a snapshot entry to a status badge (design.md §7.1 labels). */
+function statusBadge(e: ProviderSnapshotEntry): { label: string; cls: string } {
+  if (e.status === 'loading') return { label: 'Loading', cls: 'bg-muted text-muted-foreground' };
+  if (!e.enabled) return { label: 'Disabled', cls: 'bg-muted text-muted-foreground' };
+  if (e.status === 'ready')
+    return { label: 'Available', cls: 'bg-green-500/15 text-green-600 dark:text-green-400' };
+  if (e.status === 'error')
+    return { label: 'Error', cls: 'bg-red-500/15 text-red-600 dark:text-red-400' };
+  if (!e.installed)
+    return { label: 'Not installed', cls: 'bg-amber-500/15 text-amber-600 dark:text-amber-400' };
+  return { label: 'Unavailable', cls: 'bg-muted text-muted-foreground' };
+}
+
+/**
+ * v2.3 — provider management as a Settings tab section (design.md §7.1). Lists
+ * ALL registered providers (including the disabled/unavailable ones the composer
+ * picker hides). Per row: label + model count, status badge, per-id refresh,
+ * diagnostic, and an enable/disable toggle. Native boocode is always-on.
+ *
+ * Uses the home-cwd snapshot (no project arg) — provider management is global,
+ * not per-project (design.md §4.5).
+ */
+export function ProvidersSettings() {
+  const allEntries = useProviderSnapshot();
+  const [config, setConfig] = useState<CoderProvidersFile | null>(null);
+  const [busyId, setBusyId] = useState<string | null>(null);
+  const [error, setError] = useState<string | null>(null);
+  const [addOpen, setAddOpen] = useState(false);
+  const [diagId, setDiagId] = useState<string | null>(null);
+  const [diagText, setDiagText] = useState<string | null>(null);
+
+  // The raw config is needed to preserve a provider's FULL override when
+  // toggling: the PATCH replaces an id's override wholesale, so a bare
+  // { enabled } would wipe a custom ACP provider's command/label.
+  useEffect(() => {
+    api.coder
+      .getProvidersConfig()
+      .then(setConfig)
+      .catch(() => setConfig({ providers: {} }));
+  }, []);
+
+  // While any entry is loading, refetch until terminal (capped, no WS frame).
+  const pollsRef = useRef(0);
+  useEffect(() => {
+    const anyLoading = allEntries?.some((e) => e.status === 'loading') ?? false;
+    if (!anyLoading) {
+      pollsRef.current = 0;
+      return;
+    }
+    if (pollsRef.current >= 10) return;
+    const t = setTimeout(() => {
+      pollsRef.current += 1;
+      void refreshProviderSnapshot();
+    }, 2000);
+    return () => clearTimeout(t);
+  }, [allEntries]);
+
+  async function toggle(e: ProviderSnapshotEntry): Promise<void> {
+    setBusyId(e.name);
+    setError(null);
+    try {
+      const existing: ProviderOverride = config?.providers[e.name] ?? {};
+      const resp = await api.coder.patchProvidersConfig({
+        providers: { [e.name]: { ...existing, enabled: !e.enabled } },
+      });
+      setConfig({ providers: resp.providers });
+      await refreshProviderSnapshot();
+    } catch (err) {
+      setError(err instanceof Error ? err.message : 'failed to update provider');
+    } finally {
+      setBusyId(null);
+    }
+  }
+
+  async function refreshOne(id: string): Promise<void> {
+    setBusyId(id);
+    setError(null);
+    try {
+      await api.coder.refreshProviders([id]);
+      await refreshProviderSnapshot();
+    } catch (err) {
+      setError(err instanceof Error ? err.message : 'failed to refresh');
+    } finally {
+      setBusyId(null);
+    }
+  }
+
+  async function openDiagnostic(id: string): Promise<void> {
+    if (diagId === id) {
+      setDiagId(null);
+      setDiagText(null);
+      return;
+    }
+    setDiagId(id);
+    setDiagText('Loading…');
+    try {
+      const { diagnostic } = await api.coder.getProviderDiagnostic(id);
+      setDiagText(diagnostic);
+    } catch (err) {
+      setDiagText(err instanceof Error ? err.message : 'failed to load diagnostic');
+    }
+  }
+
+  const entries = allEntries ?? [];
+
+  return (
+    <div className="space-y-3">
+      <div className="flex items-center justify-between gap-2">
+        <p className="text-xs text-muted-foreground">
+          Enable, disable, refresh, or add coding agents. Disabled and unavailable providers are
+          hidden from the composer picker but managed here.
+        </p>
+        <Button variant="outline" size="sm" onClick={() => setAddOpen(true)} className="shrink-0">
+          <Plus className="size-3.5" /> Add provider
+        </Button>
+      </div>
+
+      <div className="rounded-md border divide-y">
+        {allEntries === null && (
+          <div className="px-3 py-2 text-sm text-muted-foreground">Loading…</div>
+        )}
+        {entries.map((e) => {
+          const badge = statusBadge(e);
+          const isNative = e.transport === 'native';
+          const busy = busyId === e.name;
+          return (
+            <div key={e.name} className="px-3 py-2.5">
+              <div className="flex items-center gap-2">
+                <div className="min-w-0 flex-1">
+                  <div className="text-sm font-medium truncate">{e.label}</div>
+                  <div className="text-xs text-muted-foreground">
+                    {e.models.length} model{e.models.length === 1 ? '' : 's'}
+                  </div>
+                </div>
+                <span
+                  className={cn(
+                    'inline-flex items-center rounded px-1.5 py-0.5 text-[11px] font-medium',
+                    badge.cls,
+                  )}
+                >
+                  {e.status === 'loading' && <Loader2 className="size-3 mr-1 animate-spin" />}
+                  {badge.label}
+                </span>
+                <button
+                  type="button"
+                  onClick={() => void refreshOne(e.name)}
+                  disabled={busy}
+                  className="inline-flex items-center justify-center size-7 rounded text-muted-foreground hover:text-foreground disabled:opacity-40"
+                  aria-label={`Refresh ${e.label}`}
+                  title="Refresh"
+                >
+                  <RefreshCw className={cn('size-3.5', busy && 'animate-spin')} />
+                </button>
+                <button
+                  type="button"
+                  onClick={() => void openDiagnostic(e.name)}
+                  className="inline-flex items-center justify-center size-7 rounded text-muted-foreground hover:text-foreground"
+                  aria-label={`Diagnostic for ${e.label}`}
+                  title="Diagnostic"
+                >
+                  <Stethoscope className="size-3.5" />
+                </button>
+                {isNative ? (
+                  <span className="text-[11px] text-muted-foreground w-14 text-center">
+                    Always on
+                  </span>
+                ) : (
+                  <button
+                    type="button"
+                    role="switch"
+                    aria-checked={e.enabled}
+                    disabled={busy}
+                    onClick={() => void toggle(e)}
+                    className={cn(
+                      'relative inline-flex h-5 w-9 shrink-0 items-center rounded-full transition-colors disabled:opacity-40',
+                      e.enabled ? 'bg-primary' : 'bg-muted-foreground/30',
+                    )}
+                    aria-label={`${e.enabled ? 'Disable' : 'Enable'} ${e.label}`}
+                    title={e.enabled ? 'Enabled — click to disable' : 'Disabled — click to enable'}
+                  >
+                    <span
+                      className={cn(
+                        'inline-block size-4 rounded-full bg-background transition-transform',
+                        e.enabled ? 'translate-x-4' : 'translate-x-0.5',
+                      )}
+                    />
+                  </button>
+                )}
+              </div>
+              {diagId === e.name && (
+                <pre className="mt-2 max-h-48 overflow-auto rounded bg-muted/50 p-2 text-[11px] font-mono whitespace-pre-wrap">
+                  {diagText}
+                </pre>
+              )}
+            </div>
+          );
+        })}
+      </div>
+
+      {error && <div className="text-sm text-destructive">{error}</div>}
+
+      <AddProviderModal
+        open={addOpen}
+        onOpenChange={setAddOpen}
+        onAdded={() => void refreshProviderSnapshot()}
+      />
+    </div>
+  );
+}

apps/web/src/components/panes/CoderPane.tsx

@@ -4,10 +4,11 @@
 // WS: /api/coder/ws/sessions/:id (Vite dev proxies to :9502).
 
 import { useCallback, useEffect, useMemo, useRef, useState } from 'react';
-import { Code, Check, X, RefreshCw } from 'lucide-react';
+import { Code, Check, X, RefreshCw, Terminal, Puzzle, Sparkles } from 'lucide-react';
 import { AgentComposerBar } from '@/components/AgentComposerBar';
 import { PermissionCard } from '@/components/PermissionCard';
 import { ChatInput } from '@/components/ChatInput';
+import type { SlashCommandGroup } from '@/components/SlashCommandPicker';
 import { api } from '@/api/client';
 import type { AgentSessionConfig, PermissionPrompt, AgentCommand } from '@/api/types';
 import { useSkills } from '@/hooks/useSkills';
@@ -510,6 +511,50 @@ export function CoderPane({
     [displayedCommands],
   );
 
+  // v2.5.9: segmented slash menu — the active agent's commands first, then
+  // BooCoder skills. boocode has no separate "commands" group (it IS native),
+  // so it shows only Skills. Empty groups are dropped.
+  const agentCommands = useMemo(
+    () =>
+      agentConfig.provider === 'boocode'
+        ? []
+        : mergeCommandsByName(providerCommands, liveTaskCommands),
+    [agentConfig.provider, providerCommands, liveTaskCommands],
+  );
+  const skillItems = useMemo(
+    () => skills.map((s) => ({ name: s.name, description: s.description })),
+    [skills],
+  );
+  const slashGroups = useMemo(() => {
+    const groups: SlashCommandGroup[] = [];
+    // Split the active agent's set: native/CLI commands vs plugin skills, each
+    // with its own icon. BooCoder skills always come last.
+    const agentCmds = agentCommands.filter((c) => c.kind !== 'skill');
+    const agentSkills = agentCommands.filter((c) => c.kind === 'skill');
+    if (agentCmds.length > 0) {
+      groups.push({
+        label: `${agentConfig.provider} commands`,
+        items: agentCmds,
+        icon: <Terminal className="size-3 shrink-0" />,
+      });
+    }
+    if (agentSkills.length > 0) {
+      groups.push({
+        label: `${agentConfig.provider} skills`,
+        items: agentSkills,
+        icon: <Puzzle className="size-3 shrink-0" />,
+      });
+    }
+    if (skillItems.length > 0) {
+      groups.push({
+        label: 'BooCoder skills',
+        items: skillItems,
+        icon: <Sparkles className="size-3 shrink-0" />,
+      });
+    }
+    return groups;
+  }, [agentCommands, skillItems, agentConfig.provider]);
+
   const { messages, setMessages, connected, loadMessages } = useCoderMessages(sessionId, chatId, {
     onConnectedChange,
     onPermissionRequested: (prompt) => {
@@ -736,19 +781,35 @@ export function CoderPane({
 
   const handleChatInputSlash = useCallback(async (skillName: string, userMessage: string) => {
     if (!chatId) return;
-    if (agentConfig.provider === 'boocode' && skillsByName.has(skillName)) {
+    // Only BooCoder skills route here; an agent's own commands (not skills) fall
-      setSending(true);
+    // through to a literal send in ChatInput. Skills run under the active
-      setPermissionPrompt(null);
+    // provider: boocode → native inference; external → body injected into a task.
-      setLiveTaskCommands([]);
+    if (!skillsByName.has(skillName)) return;
-      try {
+    setSending(true);
-        await api.coder.skillInvoke(sessionId, paneId, skillName, userMessage.length > 0 ? userMessage : null);
+    setPermissionPrompt(null);
-      } catch (err) {
+    setLiveTaskCommands([]);
-        toast.error(err instanceof Error ? err.message : 'skill invocation failed');
+    try {
-      } finally {
+      const data = await api.coder.skillInvoke(
-        setSending(false);
+        sessionId,
-      }
+        paneId,
+        skillName,
+        userMessage.length > 0 ? userMessage : null,
+        agentConfig.provider !== 'boocode'
+          ? {
+              provider: agentConfig.provider,
+              model: agentConfig.model || undefined,
+              mode_id: agentConfig.modeId ?? undefined,
+              thinking_option_id: agentConfig.thinkingOptionId ?? undefined,
+            }
+          : undefined,
+      );
+      if (data.task_id) setActiveTaskId(data.task_id);
+    } catch (err) {
+      toast.error(err instanceof Error ? err.message : 'skill invocation failed');
+    } finally {
+      setSending(false);
     }
-  }, [chatId, sessionId, paneId, agentConfig.provider, skillsByName]);
+  }, [chatId, sessionId, paneId, agentConfig, skillsByName]);
 
   return (
     <div className="flex flex-col h-full bg-background">
@@ -810,6 +871,7 @@ export function CoderPane({
           projectId={projectPath ?? ''}
           onSend={handleChatInputSend}
           onSlashCommand={handleChatInputSlash}
+          slashGroups={slashGroups}
           chatId={chatId ?? undefined}
           chatLabel="BooCode"
           messages={messages as unknown as import('@/api/types').Message[]}

apps/web/src/components/panes/SettingsPane.tsx

@@ -15,9 +15,10 @@ import {
 } from '@/components/ui/dialog';
 import { ModelPicker } from '@/components/ModelPicker';
 import { ThemePicker } from '@/components/ThemePicker';
+import { ProvidersSettings } from '@/components/coder/ProvidersSettings';
 import { cn } from '@/lib/utils';
 
-type Section = 'session' | 'project' | 'theme';
+type Section = 'session' | 'project' | 'theme' | 'providers';
 
 interface Props {
   session: Session;
@@ -73,7 +74,7 @@ export function SettingsPane({ session, project, maximized, onToggleMaximize, on
     <div className="flex flex-col h-full min-h-0">
       <div className="flex items-center gap-2 border-b border-border bg-muted/20 px-3 py-1.5 shrink-0">
         <div className="flex items-center gap-1 flex-1 min-w-0">
-          {(['session', 'project', 'theme'] as const).map((s) => (
+          {(['session', 'project', 'theme', 'providers'] as const).map((s) => (
             <button
               key={s}
               type="button"
@@ -116,6 +117,7 @@ export function SettingsPane({ session, project, maximized, onToggleMaximize, on
           {activeSection === 'session' && <SessionSection session={session} project={project} />}
           {activeSection === 'project' && <ProjectSection project={project} />}
           {activeSection === 'theme' && <ThemePicker />}
+          {activeSection === 'providers' && <ProvidersSettings />}
         </div>
       </div>
     </div>

apps/web/src/data/acp-provider-catalog.ts

@@ -0,0 +1,83 @@
+import type { ProviderConfigPatch } from '@/api/types';
+
+/**
+ * v2.3 Phase 5 (design.md §7.3) — a SMALL curated catalog of ACP coding agents
+ * the user might register. We deliberately do NOT port Paseo's 30+ entry list.
+ *
+ * Non-goal: we never install anything. Each entry is a manual-install hint
+ * (`installUrl` / `installCmd`) plus the config `command` that gets written into
+ * `/data/coder-providers.json`. The user installs the CLI themselves; until the
+ * binary is on PATH the provider shows as "Not installed". Commands are
+ * editable after adding — versions are aliased/untrimmed on purpose; pin on your
+ * own host once verified.
+ */
+export interface AcpCatalogEntry {
+  id: string;
+  label: string;
+  description: string;
+  /** Config command written verbatim into providers[id].command: [binary, ...args]. */
+  command: [string, ...string[]];
+  /** Where to install the CLI manually — we LINK, never install. */
+  installUrl: string;
+  /** Optional suggested install command, shown as a copyable hint. */
+  installCmd?: string;
+}
+
+export const ACP_PROVIDER_CATALOG: AcpCatalogEntry[] = [
+  {
+    id: 'amp-acp',
+    label: 'Amp',
+    description: 'Sourcegraph Amp — agentic coding CLI with an ACP bridge.',
+    command: ['amp-acp'],
+    installUrl: 'https://ampcode.com/',
+    installCmd: 'npm i -g @sourcegraph/amp',
+  },
+  {
+    id: 'gemini',
+    label: 'Gemini CLI',
+    description: 'Google Gemini CLI in ACP mode (--experimental-acp).',
+    command: ['gemini', '--experimental-acp'],
+    installUrl: 'https://github.com/google-gemini/gemini-cli',
+    installCmd: 'npm i -g @google/gemini-cli',
+  },
+  {
+    id: 'cline',
+    label: 'Cline',
+    description: 'Cline coding agent over ACP (run via npx).',
+    command: ['npx', '-y', 'cline', '--acp'],
+    installUrl: 'https://cline.bot/',
+  },
+  {
+    id: 'claude-code-acp',
+    label: 'Claude Code (ACP)',
+    description: "Zed's ACP adapter for Claude Code — distinct from the built-in PTY claude provider.",
+    command: ['npx', '-y', '@zed-industries/claude-code-acp'],
+    installUrl: 'https://github.com/zed-industries/claude-code-acp',
+  },
+  {
+    id: 'pi-acp',
+    label: 'Pi',
+    description: 'Example custom ACP entry — build the binary from source, then edit the command.',
+    command: ['pi-acp'],
+    installUrl: 'https://agentclientprotocol.com/',
+  },
+];
+
+/**
+ * Build the PATCH body that registers a catalog entry: a single-id partial
+ * providers map with the custom-ACP override (extends:'acp' + label + command),
+ * enabled. Sent to PATCH /api/providers/config (then refreshProviders([id])).
+ */
+export function buildAcpProviderConfigPatch(entry: AcpCatalogEntry): ProviderConfigPatch {
+  return {
+    providers: {
+      [entry.id]: {
+        extends: 'acp',
+        label: entry.label,
+        description: entry.description,
+        command: entry.command,
+        enabled: true,
+      },
+    },
+  };
+}

apps/web/src/hooks/useSessionChats.ts

@@ -24,6 +24,7 @@ export interface UseSessionChatsResult {
   deleteChat: (chatId: string) => Promise<void>;
   renameChat: (chatId: string, name: string) => Promise<void>;
   handleLandingSend: (paneIdx: number, content: string) => Promise<void>;
+  handleLandingSkill: (paneIdx: number, skillName: string, userMessage: string | null) => Promise<void>;
 }
 
 export function useSessionChats(
@@ -166,6 +167,25 @@ export function useSessionChats(
     }
   }, [sessionId]);
 
+  // Slash-command equivalent of handleLandingSend: the initial (landing) chat
+  // must create the chat AND assign it to the pane (openChatInPane) before
+  // invoking the skill, so the pane transitions to ChatPane and subscribes to
+  // the chat's stream. Skipping the assignment left the pane stuck on the
+  // landing page while the skill ran invisibly (and could blank the pane).
+  const handleLandingSkill = useCallback(
+    async (paneIdx: number, skillName: string, userMessage: string | null) => {
+      try {
+        const chat = await api.chats.create(sessionId);
+        setChats((prev) => (prev.some((c) => c.id === chat.id) ? prev : [chat, ...prev]));
+        openChatInPaneRef.current(paneIdx, chat.id);
+        await api.chats.skillInvoke(chat.id, skillName, userMessage);
+      } catch (err) {
+        toast.error(err instanceof Error ? err.message : `/${skillName} failed`);
+      }
+    },
+    [sessionId],
+  );
+
   return {
     chats,
     setChats,
@@ -175,5 +195,6 @@ export function useSessionChats(
     deleteChat,
     renameChat,
     handleLandingSend,
+    handleLandingSkill,
   };
 }

apps/web/src/hooks/useWorkspacePanes.ts

@@ -32,6 +32,21 @@ function chatPane(chatId: string): WorkspacePane {
   return { id: generateId(), kind: 'chat', chatId, chatIds: [chatId], activeChatIdx: 0 };
 }
 
+interface ClosedPaneEntry {
+  kind: WorkspacePane['kind'];
+  chatIds: string[];
+  activeChatIdx: number;
+}
+const MAX_CLOSED = 10;
+const closedPaneStack: ClosedPaneEntry[] = [];
+
+function pushClosed(pane: WorkspacePane): void {
+  if (pane.kind === 'empty' || pane.kind === 'settings') return;
+  if (pane.chatIds.length === 0) return;
+  closedPaneStack.push({ kind: pane.kind, chatIds: [...pane.chatIds], activeChatIdx: pane.activeChatIdx });
+  if (closedPaneStack.length > MAX_CLOSED) closedPaneStack.shift();
+}
+
 function chatNameForPaneKind(kind: 'coder' | 'terminal'): string {
   return kind === 'coder' ? 'BooCoder' : 'Terminal';
 }
@@ -50,8 +65,8 @@ export function activePaneChatId(pane: WorkspacePane): string | undefined {
 // v1.9: settings pane factory. No chats, no state beyond identity — the
 // SettingsPane component renders Session/Project sections from the
 // surrounding session/project.
-function settingsPane(): WorkspacePane {
+function settingsPane(id: string = generateId()): WorkspacePane {
-  return { id: generateId(), kind: 'settings', chatIds: [], activeChatIdx: -1 };
+  return { id, kind: 'settings', chatIds: [], activeChatIdx: -1 };
 }
 
 // v1.14.x-html-artifact-panes: artifact pane factories. Payload travels with
@@ -135,8 +150,10 @@ export interface UseWorkspacePanesResult {
   // Open-on-first-click, close-on-second-click. Singleton — settings panes
   // don't count toward MAX_PANES. Closing the only remaining pane (edge case)
   // falls back to an empty pane to preserve the "always one pane" invariant.
-  toggleSettingsPane: () => void;
+  toggleSettingsPane: () => string | null;
   removePane: (idx: number) => void;
+  reopenPane: () => void;
+  hasClosedPanes: boolean;
   removeChatFromPanes: (chatId: string) => void;
   initializeFirstChatIfEmpty: (chatId: string) => void;
   validatePanes: (validChatIds: Set<string>) => void;
@@ -391,6 +408,14 @@ export function useWorkspacePanes(sessionId: string): UseWorkspacePanesResult {
       const pane = next[paneIdx]!;
       const nextIds = pane.chatIds.filter((id) => id !== chatId);
       if (nextIds.length === 0) {
+        if (next.length > 1) {
+          // Last tab closed and other panes exist — remove the whole pane
+          // instead of leaving an orphaned empty panel.
+          pushClosed(pane); setHasClosedPanes(true);
+          const spliced = next.filter((_, i) => i !== paneIdx);
+          setActivePaneIdx((ai) => Math.min(ai, spliced.length - 1));
+          return spliced;
+        }
         next[paneIdx] = { ...pane, kind: 'empty', chatId: undefined, chatIds: [], activeChatIdx: -1 };
       } else {
         const nextActiveIdx = Math.min(pane.activeChatIdx, nextIds.length - 1);
@@ -492,14 +517,21 @@ export function useWorkspacePanes(sessionId: string): UseWorkspacePanesResult {
     return success ? newPaneId : null;
   }, [seedPaneChat]);
 
-  const toggleSettingsPane = useCallback(() => {
+  // Returns the new settings pane id when one is OPENED (so mobile callers can
+  // push ?pane= atomically — see addPaneAndSwitch), or null when it was closed.
+  // Id generated outside the updater so a strict-mode double-invoke agrees.
+  const toggleSettingsPane = useCallback((): string | null => {
+    const newPaneId = generateId();
+    let openedId: string | null = null;
     setPanes((prev) => {
       const existingIdx = prev.findIndex((p) => p.kind === 'settings');
       if (existingIdx < 0) {
-        const next = [...prev, settingsPane()];
+        const next = [...prev, settingsPane(newPaneId)];
         setActivePaneIdx(next.length - 1);
+        openedId = newPaneId;
         return next;
       }
+      openedId = null;
       if (prev.length <= 1) {
         setActivePaneIdx(0);
         return [emptyPane()];
@@ -508,6 +540,7 @@ export function useWorkspacePanes(sessionId: string): UseWorkspacePanesResult {
       setActivePaneIdx((ai) => Math.min(ai, next.length - 1));
       return next;
     });
+    return openedId;
   }, []);
 
   const removePane = useCallback((idx: number) => {
@@ -526,6 +559,7 @@ export function useWorkspacePanes(sessionId: string): UseWorkspacePanesResult {
       // The endpoint is idempotent (404 on missing session) so a strict-mode
       // double-invoke of the updater is safe.
       const removed = prev[idx];
+      if (removed) { pushClosed(removed); setHasClosedPanes(true); }
       if (removed?.kind === 'terminal') {
         api.terminals.kill(sessionId, removed.id).catch(() => { /* non-fatal */ });
       }
@@ -535,6 +569,26 @@ export function useWorkspacePanes(sessionId: string): UseWorkspacePanesResult {
     });
   }, [sessionId]);
 
+  const [hasClosedPanes, setHasClosedPanes] = useState(closedPaneStack.length > 0);
+
+  const reopenPane = useCallback(() => {
+    const entry = closedPaneStack.pop();
+    setHasClosedPanes(closedPaneStack.length > 0);
+    if (!entry) return;
+    setPanes((prev) => {
+      const restored: WorkspacePane = {
+        id: generateId(),
+        kind: entry.kind,
+        chatId: entry.chatIds[entry.activeChatIdx] ?? entry.chatIds[0],
+        chatIds: entry.chatIds,
+        activeChatIdx: Math.min(entry.activeChatIdx, entry.chatIds.length - 1),
+      };
+      const next = [...prev, restored];
+      setActivePaneIdx(next.length - 1);
+      return next;
+    });
+  }, []);
+
   // Replaces a single empty default pane with a chat pane. Used by the initial
   // chat fetch to land on the most-recent open chat if no saved pane state.
   const initializeFirstChatIfEmpty = useCallback((chatId: string) => {
@@ -664,6 +718,8 @@ export function useWorkspacePanes(sessionId: string): UseWorkspacePanesResult {
     addSplitPane,
     toggleSettingsPane,
     removePane,
+    reopenPane,
+    hasClosedPanes,
     removeChatFromPanes,
     initializeFirstChatIfEmpty,
     validatePanes,

apps/web/src/lib/slash-command.ts

@@ -18,8 +18,8 @@ export function parseSlashInput(text: string): { cmdName: string; args: string }
   return { cmdName: match[1]!, args: (match[2] ?? '').trim() };
 }
 
-export function mergeCommandsByName(...lists: SlashCommandItem[][]): SlashCommandItem[] {
+export function mergeCommandsByName<T extends SlashCommandItem>(...lists: T[][]): T[] {
-  const byName = new Map<string, SlashCommandItem>();
+  const byName = new Map<string, T>();
   for (const list of lists) {
     for (const cmd of list) {
       byName.set(cmd.name, cmd);

apps/web/src/pages/Session.tsx

@@ -123,6 +123,20 @@ function SessionInner({ sessionId }: { sessionId: string }) {
     };
   }, [sessionId]);
 
+  // v2.3: opening the settings pane on mobile must push ?pane= atomically, or
+  // the URL-sync effect below snaps activePaneIdx back to the chat pane and the
+  // settings pane never shows (same fix as addPaneAndSwitch). toggleSettingsPane
+  // returns the new pane id when it opens (null when it closes → drop ?pane= so
+  // the effect falls back to pane 0). Desktop has no URL pane state — no-op.
+  const toggleSettingsAndSync = useCallback(() => {
+    const openedId = panesHook.toggleSettingsPane();
+    if (!isMobile) return;
+    const params = new URLSearchParams(location.search);
+    if (openedId) params.set('pane', openedId);
+    else params.delete('pane');
+    navigate(`${location.pathname}?${params.toString()}`);
+  }, [panesHook, isMobile, navigate, location.pathname, location.search]);
+
   useEffect(() => {
     return sessionEvents.subscribe((event) => {
       if (event.type === 'session_renamed' && event.session_id === sessionId) {
@@ -156,10 +170,10 @@ function SessionInner({ sessionId }: { sessionId: string }) {
       // Sidebar Settings button broadcasts this when a session is mounted;
       // toggleSettingsPane opens on first click, closes on second.
       if (event.type === 'open_settings_pane') {
-        panesHook.toggleSettingsPane();
+        toggleSettingsAndSync();
       }
     });
-  }, [sessionId, editingName, navigate, project, panesHook]);
+  }, [sessionId, editingName, navigate, project, toggleSettingsAndSync]);
 
   // v1.8: URL ?pane= sync (mobile only). Lifted from Workspace.tsx so
   // MobileTabSwitcher's onSwitchPane can push the same URL state and the

data/coder-providers.example.json

@@ -0,0 +1,12 @@
+{
+  "providers": {
+    "goose": { "enabled": false },
+    "amp-acp": {
+      "extends": "acp",
+      "label": "Amp",
+      "description": "ACP wrapper for Amp",
+      "command": ["amp-acp"],
+      "enabled": true
+    }
+  }
+}

data/coder-providers.json

@@ -1,3 +0,0 @@
-{
-  "providers": {}
-}

data/skills/boocode/systematic-debugging/SKILL.md

@@ -0,0 +1,61 @@
+---
+name: systematic-debugging
+description: Guided root-cause debugging. Use when encountering any bug, test failure, unexpected behavior, or performance problem. Enforces investigation before fixes.
+---
+
+# Systematic Debugging
+
+No fixes without root cause. Symptom fixes mask real bugs and waste time.
+
+## The Rule
+
+Complete Phase 1 before proposing ANY fix. If you haven't investigated, you cannot fix.
+
+## Phase 1: Root Cause Investigation
+
+1. **Read error messages carefully.** Stack traces, line numbers, error codes. Don't skip past them.
+2. **Reproduce consistently.** Exact steps, every time. If not reproducible, gather more data instead of guessing.
+3. **Check recent changes.** Git diff, recent commits, new deps, config changes, env differences.
+4. **Trace data flow.** Where does the bad value originate? Trace backward through the call stack to the source. Fix at the source, not the symptom.
+5. **Multi-component systems:** Before fixing, add diagnostic logging at each component boundary (what enters, what exits). Run once to locate the failing layer, THEN investigate that layer.
+
+## Phase 2: Pattern Analysis
+
+1. Find working examples of similar code in the same codebase.
+2. Compare working vs broken — list every difference.
+3. If implementing a pattern, read the reference implementation completely, not skimmed.
+4. Understand all dependencies, config, and assumptions.
+
+## Phase 3: Hypothesis and Testing
+
+1. State one hypothesis clearly: "X is the root cause because Y."
+2. Make the smallest possible change to test it. One variable at a time.
+3. If it didn't work, form a NEW hypothesis. Don't stack more fixes on top.
+4. After 3 failed fixes: STOP. Question the architecture, not the symptoms.
+
+## Phase 4: Implementation
+
+1. Create a failing test case first (simplest reproduction).
+2. Implement a single fix addressing the root cause.
+3. Verify: test passes, no regressions, issue actually resolved.
+4. If the fix doesn't work and you've tried 3+: the problem is architectural. Discuss before attempting more.
+
+## Red Flags — STOP and return to Phase 1
+
+- "Quick fix for now, investigate later"
+- "Just try changing X and see"
+- "I don't fully understand but this might work"
+- "One more fix attempt" after 2+ failures
+- Proposing solutions before tracing data flow
+- Each fix reveals a new problem in a different place
+
+## Apply This Skill
+
+Use these tools to investigate before proposing changes:
+
+- `view_file` to read error sites and suspect code paths
+- `grep` to find all callers / references to the failing function
+- `find_files` to locate related config, test fixtures, schema
+- `list_dir` to understand the module layout around the bug
+
+Report your Phase 1 findings (what you observed, what you traced, what you ruled out) before moving to Phase 3.

docs/DEFERRED-WORK.md

@@ -2,7 +2,7 @@
 
 This document describes work intentionally **not** shipped in the 2026-05-26 stale/simplify batch. Each item needs a product or architecture decision before implementation. See also [`STALE-DEPRECATED.md`](./STALE-DEPRECATED.md) for what was resolved in that batch.
 
-Last updated: 2026-05-26
+Last updated: 2026-05-29
 
 ---
 
@@ -11,7 +11,7 @@ Last updated: 2026-05-26
 | Item | Category | User impact | Effort | Risk if left alone |
 |------|----------|-------------|--------|-------------------|
 | Task cancel → abort ACP/PTY child | Correctness / UX | High — Stop does not kill external agents | Medium | Zombie processes, stuck `running` tasks, orphaned worktrees |
-| Skip ACP cold probe when DB fresh | Performance | Medium — composer open can stall 5–30s on cache miss | Medium (v2.3 batch) | Slow provider picker; repeated ACP spawns on every snapshot rebuild |
+| Skip ACP cold probe when DB fresh | Performance | Medium — composer open can stall 5–30s on cache miss | ✅ Shipped (v2.3, Phase 2) | Resolved — `PROVIDER_PROBE_TTL_MS` TTL gate live |
 | Unified `packages/types` | Maintainability | Low (dev-only) | Medium–High | Type drift between server, coder, web |
 | Large file splits | Maintainability | None directly | Medium per file | Harder reviews, merge conflicts |
 | Retire `apps/coder/web/` fallback SPA | Scope / ops | Low — Sam uses CoderPane | Medium | Dual UI maintenance, divergent API client |
@@ -111,7 +111,7 @@ There is also **no frontend** calling task cancel today (`grep` across `apps/web
 
 ## 2. Skip ACP cold probe when DB models are fresh
 
-**Status:** Planned — [`openspec/changes/v2-3-provider-lifecycle/`](../openspec/changes/v2-3-provider-lifecycle/proposal.md). **Not shipped** (no `v2.3` tag; all tasks unchecked).
+**Status:** ✅ **ADDRESSED** in v2.3 (phases 1–5: `v2.5.4-provider-lifecycle-phase1` … `v2.5.12-provider-lifecycle-phase4`, plus the phase-5 settings UI + picker filter). The `PROVIDER_PROBE_TTL_MS` (default 24h) gate on `available_agents.last_probed_at` is live — the tier-2 cold ACP probe runs only on `force` (`POST /api/providers/refresh`), TTL staleness, or empty DB models; otherwise the snapshot serves cached models. See [`openspec/changes/v2-3-provider-lifecycle/`](../openspec/changes/v2-3-provider-lifecycle/proposal.md). The original (v2.2) behavior below is kept for history.
 
 ### Current behavior (v2.2)
 
@@ -140,12 +140,21 @@ See [`design.md`](../openspec/changes/v2-3-provider-lifecycle/design.md):
 
 v2.2 shipped the snapshot wire shape and ACP dispatch stack. Lifecycle semantics (config registry, enable/disable, probe TTL, settings UI) were scoped as the follow-on **v2.3** batch to avoid mixing two large behavior changes in one tag.
 
-### Acceptance criteria (when v2.3 ships)
+### Acceptance criteria — met
 
-- Second `GET /api/providers/snapshot` within TTL does not invoke `probeAcpProvider` (mock assert in tests)
+- Second `GET /api/providers/snapshot` within TTL does not invoke `probeAcpProvider` (mock assert in `provider-snapshot.test.ts`)
-- Disabled provider visible in settings, absent from composer
+- Disabled provider visible in settings (Providers tab), absent from composer
 - Explicit refresh repopulates models; warm open is sub-second
 
+### Still deferred (Tier-2 follow-ups, not shipped in v2.3)
+
+These were explicitly scoped out of v2.3 (see `design.md` §11) and remain open:
+
+- **`provider_snapshot_updated` WS frame** — the loading state uses a capped client poll / one-shot refetch instead of a server-pushed frame (design §4.4, §11; tasks O.1).
+- **`available_agents.enabled` DB column** — `enabled` is read from the in-memory resolved registry only; no DB mirror, so settings state after a coder restart re-derives from the JSON config rather than the DB (design §3.3; tasks O.2).
+- **Single-source-of-truth shared types package** — the provider snapshot types are duplicated across `apps/coder/.../provider-types.ts` and `apps/web/src/api/types.ts`, guarded by the text-identity `provider-types-parity.test.ts` rather than a shared package (see §3 below).
+- **MCP `list_providers` / `inspect_provider` tools** — provider introspection over MCP is not wired (design §11).
+
 ---
 
 ## 3. Unified `packages/types` for provider snapshot JSON

pnpm-lock.yaml

@@ -63,6 +63,9 @@ importers:
       '@modelcontextprotocol/sdk':
         specifier: ^1.29.0
         version: 1.29.0(zod@3.25.76)
+      '@opencode-ai/sdk':
+        specifier: ~1.15.0
+        version: 1.15.12
       fastify:
         specifier: ^4.28.1
         version: 4.29.1
@@ -920,6 +923,9 @@ packages:
   '@open-draft/until@2.1.0':
     resolution: {integrity: sha512-U69T3ItWHvLwGg5eJ0n3I62nWuE6ilHlmz7zM0npLBRvPRd7e6NYmg54vvRtP5mZG7kZqZCFVdsTWo7BPtBujg==}
 
+  '@opencode-ai/sdk@1.15.12':
+    resolution: {integrity: sha512-lOaBNX93dkakZe6C42ttX1bkSx3K2c6+Yv+w8Qv02v5rPlu1vCXbmdfYDh9/bw+oq+NKPSaBm9d6kPA19hA5Lg==}
+
   '@opentelemetry/api@1.9.1':
     resolution: {integrity: sha512-gLyJlPHPZYdAk1JENA9LeHejZe1Ti77/pTeFm/nMXmQH/HFZlcS/O2XJB+L8fkbrNSqhdtlvjBVjxwUYanNH5Q==}
     engines: {node: '>=8.0.0'}
@@ -4702,6 +4708,10 @@ snapshots:
 
   '@open-draft/until@2.1.0': {}
 
+  '@opencode-ai/sdk@1.15.12':
+    dependencies:
+      cross-spawn: 7.0.6
+
   '@opentelemetry/api@1.9.1': {}
 
   '@pinojs/redact@0.4.0': {}