Scoped half of boocode_code_review_v2 §1 #10 — publish the agent status
BooCoder already observes (the config-injection notify-hook is the documented
follow-on, clean-room from superset ELv2).
- agent_status_updated WS frame (working|blocked|idle|error), server+web parity.
- Published from the dispatcher's turn boundaries (warm-acp/opencode/sdk/pty:
working at start, idle/error at end) + the permission flow (blocked/working).
Best-effort, never breaks a turn.
- Clean-room normalizeAgentEvent helper (superset's vendor-event -> Start/blocked
/Stop collapse, event names as facts) + 25 tests — reused by the follow-on.
- AgentComposerBar status dot (distinct from the WS-liveness dot), tracked per
(chat,agent) by a useAgentStatus map in CoderPane.
Built by 2 parallel agents vs a pinned frame contract. Server 545 + coder 294
tests passing (25 new); web tsc + builds clean; ws-frames parity green. Clears
the actionable review backlog (#1/#3/#4/#6-#12). Builds on v2.7.5.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Lands the lean-SDK direction (boocode_code_review_v2 §1 #9) behind a flag.
Adds @anthropic-ai/claude-agent-sdk@0.3.159 (Commercial Terms, runtime dep).
- PostgresSessionStore: clean-room impl of the SDK's real SessionStore type
over a new claude_session_entries table. Typechecks against the SDK type;
8 DB-integration tests.
- ClaudeSdkBackend (implements AgentBackend): one warm query() per (chat,claude)
in streaming-input mode via a pushable async-iterable pump, sessionStore +
resume continuity, pure mapSdkMessage->AgentEvent, session_id from init,
usage/cost onto agent_sessions (backend CHECK gains 'claude_sdk').
- Routing env-gated by CLAUDE_SDK_BACKEND (default off) -> PTY path UNCHANGED.
- Built against real SDK 0.3.159 types (install paid off: partial=stream_event
needing includePartialMessages, MessageParam, result error arm).
- Fix latent test-infra deadlock: serialize DB suites (fileParallelism:false).
Coder 269 passing default / 290 with DB; tsc clean vs SDK types; builds clean.
LIVE pump + resume + actual claude turn need a host smoke (CLAUDE_SDK_BACKEND=1
+ claude binary + auth). zod peer-dep wants ^4 (workspace 3.25). Builds on v2.7.4.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two native-inference hardening features from boocode_code_review_v2 §1 #12.
MistakeTracker: new pure mistake-tracker.ts tracks consecutive heterogeneous
tool failures (kinds surfaced per tool from tool-phase.ts). On 3 in a row the
turn loop soft-nudges (model-facing recovery guidance + mistake_recovery
sentinel + reset), then escalates to stopping the turn (cap-hit-style, Continue
affordance) on a re-trip. Complements doom-loop (identical repeats) + cap-hit.
File-provenance ledger: compaction.ts derives a deterministic ## Files Read list
from the head messages' read-tool calls and injects it into the rolling-summary
prompt so provenance survives compaction (no new table; read-only).
mistake_recovery sentinel: MessageMetadata arm (server + web) + MessageBubble
render branch. Built by 2 parallel agents. Server 545 tests passing (23 new);
build + web tsc clean. Native-inference only. Builds on v2.7.3.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Three small wins from boocode_code_review_v2 §1 #11/#7/#8.
#11 sampling knobs: top_n_sigma + dry_* family as first-class Agent fields,
threaded into the request body via providerOptions.openaiCompatible. Fixes a
latent bug — top_k (rejected by the AI-SDK provider) and min_p (never passed to
streamText) were dead on the wire; both now route through the same channel.
--reasoning-budget documented in data/AGENTS.md.
#7 live PTY stream-json: new stream-json-parser.ts line-buffers qwen/claude
NDJSON and emits text/reasoning/tool frames live + persists, with a fallback to
the old opaque slice. claude gets --output-format stream-json --verbose.
#8 token UI: agent_sessions input/output_tokens/cost now flow through the route
+ type and render beside the AgentComposerBar session chip.
Built by 3 parallel agents. Server 523 + coder 245 tests passing; builds + web
tsc clean. Builds on v2.7.2. openspec sampling-streamjson-tokens.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Flagged by the automated push security review on v2.7.1.
- GET /checkpoints?chat_id= : the chat_id branch filtered by chat_id alone
(any session's chat_id read its checkpoints). Now joins chats and gates on
chats.session_id.
- restoreCheckpoint scope guard was fail-open: `cp.session_id && cp.session_id
!== sessionId` fell through on a null denormalized session_id, allowing a
cross-session restore (worktree reset + transcript trim). Now resolves the
owning session via the checkpoint's chat and denies on missing/mismatch.
- Adds a DB-integration regression for the null-session_id cross-session case.
Both scope authoritatively through chats.session_id (checkpoints.session_id is
a nullable hint). Coder suite 234 passing; 7/7 checkpoint tests (incl. the
regression) against live postgres+git; typecheck clean. Hotfix on v2.7.1.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
#3 Fuzzy patch applier: new pure fuzzy-match.ts (locateMatch, exact→trim→
unicode-canon→Levenshtein≥0.66, refuse-on-ambiguous) wired into pending_changes
applyOne/rewindOne so local-model whitespace/unicode drift in old_string no
longer loses the edit.
#4 Worktree checkpoint + conversation-trim: checkpoints table + checkpoints.ts
(shadow-commit of tracked+untracked into refs/boocode/checkpoints, hooked into
the 3 external-agent dispatcher paths) + POST restore route (reset --hard +
clean -fd -> transcript trim -> backend-session reset) + "Restore to here" UI.
Built by 3 parallel agents; DB-integration testing caught a created_at
self-deletion bug. Coder suite 234 passing; server+coder build + web tsc clean.
Builds on v2.7.0-mit. openspec write-edit-robustness.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
apps/server fire-and-forgets BooCoder's Phase-3 close hooks (new coder-notify.ts, reuses BOOCODER_URL, never-rejects) on session-delete + chat archive/archive-all/delete, so warm backends + worktrees tear down immediately (idle-evict/reaper was the backstop). 3.7: BooCoder DiffPanel shows a muted one-liner when the selected provider can't see another agent's unapplied worktree edits (pure derivation from per-change agent + current provider, no new state). 6 new server tests (coder-notify); 537 server tests pass; web+server tsc/build clean.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
WarmAcpBackend (AgentBackend) holds one persistent goose acp / qwen --acp child + ClientSideConnection + ACP session per (chat,agent); initialize+session/new once, reused across turns. Abort = session/cancel the prompt only (never kills the child); child exit -> agent_sessions.status='crashed' -> re-spawn next turn. Dispatcher routes goose/qwen chat-tab tasks to the pooled warm backend via pure shouldUseWarmBackend (needs session_id+chat_id); one-shot runExternalAgent kept as fallback for arena/MCP/new_task. handleSessionUpdate extracted to a shared pure acp-event-map.ts (one-shot path byte-identical). SDK: installed @agentclientprotocol/sdk@^0.22.1 has stable resumeSession/loadSession; resume moot in the warm hot path, deferred to Phase 3. 15 new tests (warm-acp-routing, acp-event-map); 180 coder tests pass; tsc + build clean.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
CHANGELOG + roadmap shipped record (through v2.6.8) + openspec v2-6 Phase 1-UX checked off (U.1-U.6; Smoke U pending the frontend Docker rebuild).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
DiffPanel renders a per-row agent badge (icon+label; null -> 'manual') + a 'Changes from X, Y' note when the pending set spans >1 agent. AgentComposerBar gains an optional sessionId prop -> resumed/history/new-session chip beside the Provider picker (gated, so BooChat callers are unchanged), driven by a new useAgentSessions hook (refetch on message-complete). providerIcon extracted to shared components/coder/providerIcons.tsx; api.coder gains agentSessions(sessionId); PendingChange type gains agent. web tsc clean.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
pending_changes.agent stamped at every queue site (native -> 'boocode', dispatched external -> task.agent, manual RightRail -> NULL) + flows through listPending. New GET /api/sessions/:id/agent-sessions -> [{agent,status,has_session,last_active_at}] per (chat,agent). opencode warm server consumes session.next.step.ended, accumulating input_tokens/output_tokens/cost onto agent_sessions (new idempotent columns) via a pure opencode-usage.ts mapper. Tests: agent-sessions.routes (3) + opencode-usage (6); tsc clean.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
opencode emits one trailing session.idle/error for a turn cancelled via client.session.abort(), carrying only a sessionID (no turn id). The warm-server backend settled activeTurn on that event, so after Stop + an immediate new message the orphan idle settled the NEXT turn early as success (one-click reachable since v2.6.5's Send->Stop composer).
Adds a pure per-session guard (backends/turn-guard.ts: armAbortGuard / noteTurnActivity / consumeTerminal over swallowNextTerminal) wired into opencode-server.ts: abort arms it, the next terminal is swallowed once, and a new turn's first delta self-heals so a never-arriving orphan can't strand a real turn. Test-first; 3 regression tests in turn-guard.test.ts. Paseo parallel: 1d38aac.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Capture four recurring gotchas from the panes/tabs/composer batch: the
workspace_panes WorkspaceState envelope (+ legacy-array migration on hydrate and
the union-accepting server PATCH validator); the optional ToolExecCtx
({ sql, sessionId }) 4th arg on ToolDef.execute for DB/session-aware tools
(read_tab_by_number reference); the two-schema-files-one-DB ownership split
(apps/coder owns agent_sessions/worktrees/pending_changes/available_agents) plus
the idempotent confdeltype FK-action-flip pattern; and that React StrictMode is
on, so a setState called inside another setState's updater double-fires in dev.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The composer's primary button now reflects generation state: Send when idle,
Stop while generating with an empty draft, and Queue while generating with a
draft typed (submitting queues it via the existing queue path). Stop is
click-only so a stray Enter never interrupts a run. ChatInput gains generating
+ onStop props.
BooChat: removes the separate centered "Stop generating" pill and wires
generating={streaming} + onStop={handleStop}. BooCoder: generating now keys on
sending || activeTaskId (the dispatch POST is too brief on its own), which also
fixes the queue gates that previously fired mid-run; onStop cancels the active
task via the new api.coder.cancelTask, and the input is no longer disabled while
a task runs so follow-ups can be queued.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A cohesive batch of pane/tab UX + the persisted workspace-state model (grouped
because the changes interleave across useWorkspacePanes, ChatTabBar, Workspace,
sessionEvents and the api types/client):
- Open a whole chat in a fresh pane via a new open_chat_in_new_pane event:
ChatTabBar tab context menu "Open in new pane", and MessageBubble.fork() now
lands the fork beside the original instead of replacing the active pane.
openChatInNewPane detaches the chat from any pane already holding it
(one-chat-per-pane).
- The tab-bar "+" becomes a New BooChat/BooTerm/BooCode menu (chat as a tab,
term/coder as split panes); the split button is unchanged.
- Drop the per-message "Open in pane" button (it opened a single message's
artifact) and its dead code; the artifact-pane machinery is left orphaned for
a later teardown.
- Session history: the empty/landing pane lists the session's open chats plus
archived chats (fetched separately), click to open / restore-and-open.
- Relocate-on-close: closing a chat pane moves its tabs (in order) into the
oldest chat/empty pane instead of discarding them; terminal/coder panes close
as before. Reopen strips the restored chatIds from all live panes first, so a
relocated-then-reopened pane never duplicates a tab — no stack-shape change.
- Stable global tab numbering: tabNumbers/nextTabNumber assigned on chat-pane
open, retired on close (never reused), rendered map-keyed (not positional).
- workspace_panes is now a WorkspaceState envelope { panes, tabNumbers,
nextTabNumber, closedPaneStack }; the reopen stack moved from a module-level
array into the persisted envelope so it survives reload. Hydrate/persist
normalize the legacy bare-array shape. appendClosed dedupes a value-identical
top entry to neutralize the StrictMode double-invoke of the setPanes updater.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Widen the sessions.workspace_panes JSONB from a bare WorkspacePane[] to a
WorkspaceState envelope { panes, tabNumbers, nextTabNumber, closedPaneStack }.
The PATCH validator accepts either the legacy array or the envelope (zod union)
and normalizes to a full envelope before storing, so existing array-shaped rows
migrate transparently on next write. The session_workspace_updated WS frame
schema is widened to match (kept byte-identical to the web copy; parity test
passes).
Adds read_tab_by_number, a read-only tool that resolves a session-scoped tab
number to its chat via the persisted tabNumbers map and returns that chat's
transcript (oldest-first, sentinels skipped, capped at 20k chars). Tools gain an
optional ToolExecCtx ({ sql, sessionId }) 4th param on ToolDef.execute, threaded
through executeToolCall from executeToolPhase; the param is optional so existing
filesystem tools and the apps/coder consumer stay compatible. Registered in
ALL_TOOLS + READ_ONLY_TOOL_NAMES.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
flattenToMessage now places the typed text first and appends pasted-chip
content after it with a single leading space (file/line chips remain fenced
provenance blocks after that), instead of prepending all attachments. A
leading slash command therefore stays first and the paste reads as its
continuation — `/command <pasted>` rather than `<pasted>` then the command.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Reflect shipped v2.6.1–v2.6.3 work in the deep reference. The opencode SSE
bullet now describes per-session SSE (P1.5-a) instead of the single-stream
Phase-1 limit; the agent_sessions resume bullet describes the (chat_id, agent)
re-key (P1.5-b) — chat_id CASCADEs from chats, session_id/worktree_id are
informational SET NULL, and the worktrees table supersedes the defanged
session_worktrees. Drop the stale root AGENTS.md navigation pointer (removed
in v1.12; data/AGENTS.md is the registry, not navigation). Add two
conventions: data/AGENTS.md is parsed (## headings need a --- fence, no
free-form rule sections) and the data/skills/<vendor>/ layout with the
boocode/ namespace.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The P1.5-b re-key block (cb1846c) re-adds session_id_fkey as ON DELETE
SET NULL, but the whole block is guarded on chat_id_fkey's absence. A DB
already re-keyed to (chat_id, agent) while session_id_fkey was still
ON DELETE CASCADE never re-enters that block, so applySchema leaves it at
'c' forever — diverging from the schema's stated intent, from worktree_id
(already SET NULL), and from the v2.6.3 changelog's own claim that
session_id is informational SET NULL.
Add a standalone confdeltype-guarded block (mirroring the session_worktrees
defang) that flips session_id_fkey CASCADE -> SET NULL independently of the
re-key gate. Idempotent: fires only while the FK is still 'c' — a no-op on a
fresh deploy (already 'n' from the re-key block) and on every re-run. The
live DB was converged by hand with the identical statements; \d
agent_sessions now shows session_id ... ON DELETE SET NULL.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Two portable agent-judgment skills in data/skills/boocode/, externalizing when/how Opus commits and when it isolates work in a worktree, so weaker agents (opencode build agent, BooCoder) can approximate it. committing-changes: segment by concern, stage explicitly (never git add -A), draft scope-prefix messages, present-and-STOP — commit only on explicit command, never push, identity indifferentketchup. using-worktrees: the when-to-isolate heuristic (just-create-when-clear / propose-when-ambiguous / skip), stable-base mechanics, runtime-isolation caveat — deliberately autonomous vs committing's command-gate. Each has an eval.yaml (matching improving-boocode-guidance) with a negative-trigger task. AGENTS.md gets a parser-safe preamble (the registry throws on bare ## headings) pointing at both skills.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The tab (a chat) is the context unit: two opencode tabs in one session are two independent agent contexts sharing one worktree. agent_sessions re-keys from (session_id, agent) to (chat_id, agent) — chat_id FK ON DELETE CASCADE (closing a tab ends its context); worktree_id and session_id become informational SET NULL columns. New worktrees table (one-per-session, survives session delete via session_id SET NULL) supersedes session_worktrees, which is defanged (CASCADE dropped) not yet removed. chat_id is threaded end-to-end: tasks.chat_id added, written by the coder message + skills routes from the frontend tab, read by runOpenCodeServerTask which falls back to resolve-or-create a chat for session-less creators (arena/MCP/new_task/generic) so ensureSession never gets a null key. Idempotent migration with a backfill-verify gate (0-row assertion after the test session was deleted). config_hash fingerprint logic preserved; one-worktree-per-session unchanged; runExternalAgent untouched. Column rename worktree_path -> path repointed at all five readers (server delete-guard, risk/stash endpoints, ensureSessionWorktree). Supersedes the earlier (worktree_id) draft.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The dcp tag (<dcp-message-id>mNNNN</dcp-message-id>) is streamed token-by-token, so it arrives split across SSE deltas. The existing per-chunk stripDcpTags never sees a complete tag in any single fragment, so fragments pass through and the dispatcher reassembles the tag in textChunks (persisted + shown) — and the terminal message.part.updated path that would strip the full text is suppressed by the dedup gate. Add a stateful cross-chunk stripper (dcp-strip.ts: makeDcpStreamStripper) at the dispatcher's opencode frame boundary: it emits text that cannot be part of a forming tag, holds back only a trailing partial-tag prefix (without swallowing legitimate <…> content), and flushes at turn end. Fixes both live delta frames and persisted content. 11 unit tests incl. split-at-every-boundary and the documented per-chunk-fails case. opencode path only; ACP (goose/qwen/claude) untouched.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Session worktree branches (session-<id>) never get an upstream, so the original atRisk rule (unpushed !== 0) flagged every worktree-backed session as at-risk on delete — even pristine ones — forcing a Stash/Force confirm on each. Gate the unpushed arm behind hasUpstream (unpushed !== -1) so the no-upstream sentinel can't trigger it: atRisk = dirty || unmerged > 0 || (hasUpstream && unpushed > 0). No protection is lost — any genuinely unsafe local commit also shows as unmerged > 0 — and the unpushed > 0 arm stays correct for P1.5's pushable worktree branches. unpushed is still reported (-1 = local-only) as informational. Follow-up to 3a26563.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace the single global SSE loop (scoped to the most-recently-used worktree directory) with one subscription per live opencode session, each scoped to that session's worktree dir. Two sessions in different worktrees now stream concurrently instead of the second silently dropping the first's events. Each session owns an AbortController (SessionState.sseAbort) wired into subscribe(..., {signal}); the loop reconnects, reconciles (per-session), and is torn down on closeSession/dispose by aborting the signal — which also fixes a latent Phase-1 bug where switching directories left the old runEventLoop parked forever in its for-await (zombie loops). A sessionID demux guard (eventSessionId) drops events that aren't this loop's own, so two sessions sharing a worktree (possible after P1.5-b) don't double-process each other's deltas. Removed sseRunning/sseDirectory/startEventLoop/runEventLoop/reconcileInFlight and the 'SSE directory changed' collision warning. dispatchEvent/handleUpdatedPart (translation, dedup, dcp-strip) and the watchdog are unchanged — only the subscription topology changed. SDK confirmed: @opencode-ai/sdk Event.subscribe opens an independent SSE connection per call, so N concurrent dir-scoped streams are supported. No schema/dispatcher/frontend changes; runExternalAgent untouched.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Deleting a BooChat session CASCADE-wipes its session_worktrees row, which would silently orphan uncommitted/unpushed/unmerged work in the worktree. Add a pre-DELETE gate: the server reads session_worktrees from the shared DB first (no row = chat-only session = delete immediately, zero round-trip), and for worktree-backed sessions calls a new BooCoder endpoint that runs git on the host (only the host systemd service can see /tmp/booworktrees). checkWorktreeWorkAtRisk reports dirty/unpushed/unmerged via the audited hostExec+shellEscape path; default branch is detected from refs/remotes/origin/HEAD (not the worktree's own branch), never hardcoded. Any at-risk worktree returns 409 with per-worktree RiskReport[]; force=true bypasses the check entirely. Fail-closed: coder unreachable/errored also blocks (force still escapes). The sidebar renders a block dialog distinguishing work-at-risk (Commit/Stash/Force) from couldn't-verify (Cancel/Force only); stash uses -u and re-blocks on remaining commits with an explanatory message. Commit never auto-commits — it routes the user to the session.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
- opencode is now a warm HTTP server (was "planned, unshipped").
- SSE: session.next.* event types + subscribe({directory}) requirement.
- Model strings need llama-swap/ prefix + presence in opencode.json.
- config_hash excludes ephemeral port; session FKs are ON DELETE CASCADE.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
