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.
Add top_p/top_k/min_p/presence_penalty to AGENTS.md frontmatter and thread
through inference (agents.ts parser → Agent type → stream-phase → sentinel
summaries). Null means omit from request body, preserving provider defaults.
Wire ask_user_input interactive card into both BooCoder frontends: the
CoderPane in BooChat's SPA (CoderMessageList now renders AskUserInputCard
instead of ToolCallLine for ask_user_input tool calls) and the standalone
coder SPA (MessageBubble + new AskUserInputCard + shadcn ui primitives).
Additional fixes: SessionLandingPage uses ChatInput with slash-command
support and lazy chat creation; Session.tsx hydrate-race fix for empty pane
promotion; AgentPicker wider dropdown with line-clamp; ModelPicker min-width;
Textarea converted to forwardRef; Recon agent added to AGENTS.md; codecontext
host port exposed in docker-compose.
Generalizes the v1.14.1 single-server Context7 PoC into a multi-server MCP
client registry with per-server graceful degradation. JSON config at
/data/mcp.json (bind-mounted alongside AGENTS.md) matches opencode's
mcpServers schema shape. Config file missing = no MCP (opt-in by presence).
Two transports: Streamable HTTP (remote servers like Context7) and stdio
(local subprocess servers like codecontext). Stdio spawns a persistent child
via the SDK's StdioClientTransport; shutdown hook closes all transports.
Tool prefix generalized from context7_<name> to <serverName>_<toolName> with
a toolToServer reverse map for dispatch routing. AGENTS.md tools: field now
supports glob patterns (context7_*, !web_*) via matchToolGlob — last-match-
wins with ! deny prefix. Replaces exact-match .includes() in stream-phase.ts.
refreshToolNames() in agents.ts rebuilds the DEFAULT_TOOLS snapshot after
appendMcpTools so agents without explicit tools: lists see MCP tools —
reviewer caught that the module-load-time snapshot would permanently exclude
late-registered tools.
Read-only invariant: readOnlyHint === false rejected at discovery. Result
size capped at 5MB. v1.14.1 env vars removed — superseded by config file.
Default data/mcp.json ships with Context7 disabled.
363/363 server tests passing. No schema changes, no frontend changes.
Converts the ad-hoc executeToolPhase → runAssistantTurn recursion into an
explicit while (stepNumber < effectiveCap) loop. A step is one stream-and-
tool-execute iteration; the loop terminates on non-tool finish, step-cap hit,
doom-loop, budget exhaustion, abort, or synthesis success.
MAX_STEPS = 200 hard ceiling (4x old effective limit from budget). Per-agent
steps: field in AGENTS.md frontmatter sets tighter caps (Refactorer: 5,
Architect: 20, others: unset = bounded only by MAX_STEPS). Resolution:
effectiveCap = Math.min(agent.steps ?? Infinity, MAX_STEPS).
executeToolPhase no longer recurses — returns ToolPhaseResult struct
(action: 'continue' | 'paused' | 'synthesis_done') so the caller decides
whether to continue or break. steps: 0 handled as "no tool calls allowed"
via runTextOnlyTurn (one text-only stream phase, tool calls ignored with
warn log).
Step-cap hits produce a sentinel summary (reuses cap_hit kind so
CapHitSentinel.tsx renders without frontend changes; text distinguishes
"Step limit reached" from "Tool budget exhausted"). Doom-loop check migrated
to top of loop body — same predicate, same threshold (3), break instead of
return.
step_start parts are in the schema CHECK but not emitted as message_parts —
writing before the stream phase creates a sequence-0 collision with
partsFromAssistantMessage. Structured log line emitted instead. Adversarial
review caught the collision pre-deploy.
332/332 server tests passing. No frontend changes. No schema changes.
Pattern lift from eyaltoledano/claude-task-master (MIT + Commons Clause
— pattern only, no code lift). Adds BOOCODE_TOOLS env var with three
tiers:
- core (4 tools): view_file, list_dir, grep, find_files. ~2k token
schema cost.
- standard (15 tools): core + web_search, web_fetch, git_status, all
8 codecontext_* tools. ~10k token schema cost.
- all (default; current behavior): every tool in ALL_TOOLS (20). ~21k
token schema cost.
The env var is a CEILING — narrows agent whitelists, never expands.
Default behavior unchanged when var is unset. resolveToolTier is
case-insensitive and falls back to 'all' on unknown values.
CORE_TOOL_NAMES + STANDARD_TOOL_NAMES validated at module load against
TOOLS_BY_NAME via two top-level for-loops that throw on the first
missing name. Module fails to import if a tier references a tool that
doesn't exist in the registry — catches typos and stale tier
definitions at boot rather than silently filtering valid tools out of
agent whitelists.
Wiring: agents.ts parseAgentBlock now reads BOOCODE_TOOLS from
process.env per parse, intersects with the agent's declared frontmatter
tools (or DEFAULT_TOOLS when frontmatter omits the field). Per-parse
read is fine — agents are re-parsed on the existing 60s cache TTL.
Tests: tools.test.ts grows from 1 to 10 tests. Covers resolveToolTier
across tiers/case/unknown values + the CORE-subset-of-STANDARD invariant
+ TOOLS_BY_NAME existence for both tier sets. 204/204 pass (was 195;
+9 new).
Deviation from the brief: the codecontext tools in the actual registry
have NO codecontext_* prefix (the brief's STANDARD list assumed it).
Used the actual names (get_codebase_overview, search_symbols, etc.).
Module-load validation would have failed boot with the prefixed names.
Smoke: with BOOCODE_TOOLS unset, agents return their full 12-tool
whitelists. With BOOCODE_TOOLS=core in .env + container restart, the
same agents narrow to 4 tools (find_files, grep, list_dir, view_file)
— intersection of declared whitelist ∩ core tier. Reverted after
confirmation.
CLAUDE.md updated with BOOCODE_TOOLS in the Environment section's
Optional list. .env.example gained a commented BOOCODE_TOOLS=all line
with the per-tier token-cost table.
~110 LoC across 5 files (4 modified + 1 test expansion). Under the
brief's ~30 LoC estimate for code; the test suite expansion drove
most of the growth.
Recon during planning disproved the original v1.13.7 (DB-cache) premise:
buildSystemPrompt already runs over inputs mtime-cached at the file layer
(BOOCHAT.md in system-prompt.ts:25, AGENTS.md global+per-project in
agents.ts:245), and DB scalars are byte-stable until edited. The output
is microsecond pure-string concat with no I/O. Skills aren't in the
prefix; tools live in a separate request body field alpha-sorted by
v1.13.3.
This batch closes the verification gap with instrumentation, not
implementation:
- system-prompt.ts: buildSystemPromptWithFingerprint canonical impl
computes SHA-256 over the assembled prefix, runs a per-session
Map<sessionId, lastHash> observer, emits PrefixFingerprint per call
and PrefixDrift (with field-level changed_inputs) on hash change.
buildSystemPrompt is now a thin shim returning .prompt.
- agents.ts: getAgentsMtimes accessor — cache-read only, no I/O.
- payload.ts: buildMessagesPayload takes optional log argument; when
passed, emits prefix-fingerprint (info) + prefix-drift (warn).
- turn.ts + sentinel-summaries.ts: pass ctx.log at 3 production call
sites; sentinel summaries log too so any drift across cap-hit /
doom-loop paths surfaces.
- system-prompt.test.ts: 4 new tests (byte-identical, no-drift-on-
stable, drift-fires-with-changed-inputs, cross-session-no-drift).
194/194 tests pass (was 190).
Smoke: 5 messages in a fresh session produced 7 prefix-fingerprint
logs (extras from buildMessagesPayload being called from sentinel
summary paths), all with identical prefix_hash and prefix_length=2907,
zero prefix-drift. Prefix is byte-stable in steady-state.
Decision: original system_prompt_cache DB table from the roadmap is
permanently dropped. The v1.12.0 mtime caches at the input layer plus
alpha tool ordering at the request body (v1.13.3) already address the
load-bearing cache-stability surfaces. Instrumentation stays so the
claim can be re-verified at any time.
Removed /opt/boocode/AGENTS.md (per-project override) — the project's
agents now resolve from the global /data/AGENTS.md only. Eliminates the
two-files-must-stay-in-sync footgun that surfaced during B.3
verification.
Fix: agents.ts ALL_TOOL_NAMES was a hardcoded 9-item whitelist that
silently filtered any unknown tool name from agent.tools arrays. This
caused web_search/web_fetch (v1.11.8) and the 8 codecontext tools to be
dropped at parse time. Replaced with ALL_TOOLS.map(t => t.name) for
single source of truth. Pre-existing exposure was dormant since no
builtin agent listed web_search; surfaced by adding codecontext.
- /data/skills mount (host: /opt/skills)
- skill_find, skill_use, skill_resource added to default read-only
tool set; opt-in for agents with explicit tools: whitelist
- AGENTS.md builtin agents drop explicit tools: arrays to inherit
the new default (now includes skill tools)
- POST /api/chats/:id/skill_invoke for slash-command flow
- 19 SKILL.md files seeded at /opt/skills/ across 6 source groups
Old hardcoded MAX_TOOL_LOOP_DEPTH=15 replaced by per-agent
max_tool_calls (1-100, AGENTS.md frontmatter) with defaults: 30 for
read-only-only agents, 10 for agents that include any non-read-only
tool, 15 for raw chat. When the loop hits cap, fire one final summary
call with tools disabled, stream the wrap-up into the in-flight
assistant message, then insert a system sentinel with
metadata.kind='cap_hit'. The sentinel renders an amber bubble with a
Continue button (latest sentinel only) that POSTs to a new
/api/chats/:id/continue route to extend. Hard ceiling: 3 cap-hits per
chat (2 continues max) — third sentinel reports can_continue=false.
Error frames carry a machine-readable reason code alongside human
error text. Failed messages persist the reason via
metadata.kind='error' so the bubble renders specifics on reload (WS
error frame is one-shot).
Tool call UI rewired: ToolCallLine renders inline (↳ name args
spinner/check/✗, expand-on-tap for args+result); ToolCallGroup
collapses 3+ consecutive same-tool runs into a compact card.
MessageList owns a three-pass pre-render (flatten + fold tool
results onto matching runs by id + group same-tool runs + number
sentinels). MessageBubble drops tool rendering and adds the
sentinel / error-reason branches. ToolCallCard deleted.
Roadmap follow-up logged: add explicit max_tool_calls: 30 to the 6
agents in /data/AGENTS.md and /opt/boocode/AGENTS.md post-ship for
discoverability (defaults handle behavior identically).
Builtins move out of code into /data/AGENTS.md (always-on, mounted ro
into the container); per-project AGENTS.md is now an optional override.
agents.ts merges global + project entries with project-wins-by-name and
caches per-source mtimes (60s TTL). Parser switches to per-block
try/catch and returns AgentsResponse { agents, errors[] } so one
malformed block no longer fails the file. AgentPicker shows a
non-blocking amber chip listing skipped blocks and only fires a gray
toast when zero agents loaded.
WS reconnect UX (useUserEvents + useSessionStream) now silent on the
first disconnect; createWsReconnectToast escalates to gray after 3
failures or 15 s, then to red with a Retry Now action after 60 s.
useSessionStream also gained the exponential-backoff reconnect it was
missing.
Six builtin defaults (Code Reviewer, Debugger, Refactorer, Architect,
Security Auditor, Prompt Builder) with no model field so session.model
wins. Project root AGENTS.md parsed on demand with mtime cache; when
present, only its agents are shown. sessions.agent_id resolves per turn
into effective system prompt, temperature, and a tool whitelist applied
in inference. AgentPicker mounts in the ChatInput toolbar; SettingsDrawer
agent surface deferred to Batch 7.
