apt-mcp-agent-setup 3.0.9 → 3.2.5

package/src/templates/mcp-tools.md

@@ -4,12 +4,13 @@
 
 For Figma and Stitch tool names and authentication, read `docs/agents/design-config.md` when working on design tasks.
 
-- Bootstrap: `session_bootstrap(sessionId, contextEpoch, host, capabilities)` identifies this context and reports resumable work.
-- Route: `route_request(sessionId, requestId, prompt, target, stack, explicitSkills)` returns the workflow and companion skills. Load them using `skill_load` only when needed.
-- Start: `pipeline_start(pipeline, task, workspace, acceptanceCriteria, scope, sessionId)` creates a versioned run. Use `pipeline_use(task, runId)` to resume the same run.
-- Execute: `pipeline_next(task, runId, sessionId, role, requestId)` reserves an attempt. Native host agents perform the work; the coordinator controls transitions.
-- Verify: `pipeline_verify` records authorized argv execution under the run workspace. `pipeline_checkpoint` references evidence IDs for completion and supports report, pause, resume, release and criteria actions.
-- Inspect: `pipeline_status` is read-only and does not bind sessions. `pipeline_reset` cancels without deleting history. `pipeline_prepare_merge` only prepares a report.
+- Bootstrap: `session_bootstrap(sessionId, contextEpoch, host, capabilities)` identifies this context and reports resumable work. Pass native model capability and account scope only when the host supplies them.
+- Route: `route_request(sessionId, requestId, prompt, target, stack, explicitSkills)` returns the workflow, specialist, review roles, platform verification requirements and phase-specific skills. Load current-step skills with `skill_load` and agent/workflow content with `context_load` for each new worker or context epoch.
+- Models: `model_list(sessionId, offset, limit, query)` reports inventory source, account scope, coverage and eligibility. CLI metadata is reference only. Use the model decision from `pipeline_next`; compare requested and host-observed model receipts after dispatch.
+- Start: `pipeline_start(pipeline, task, workspace, acceptanceCriteria, scope, qualityPlan, productImpact, sessionId)` creates a versioned run. New executable runs require at least one acceptance criterion string that exactly matches `qualityPlan.criteria[].expectedOutcome`. The plan uses `criteria: [{id, expectedOutcome, source, checkIds}]`, `checks: [{id, criterionIds, format, testIds, required}]`, and `reviews: [{id, required}]`. A recent user-facing web or mobile route requires its platform check and Product review. Use `pipeline_use(task, runId)` to resume the same run.
+- Execute: `pipeline_next(task, runId, sessionId, role, requestId)` reserves an attempt. Include its APT attempt markers in the native worker brief. Native host agents perform the work; the coordinator controls transitions and continues through ready parts of an approved feature without asking for approval after each internal step. Resource and quota waits remain resumable.
+- Verify: `pipeline_verify` records authorized argv execution under the run workspace. For functional acceptance, pass `checkId`, supported `reportFormat` and named test IDs in the pinned plan. Zero, skipped or unreadable tests and unrelated successful commands remain unverified. `pipeline_checkpoint` records the latest review verdict per stage and persistent finding IDs through `action=review`; resolution needs current evidence and later reviewer confirmation. A direct start without a bound route needs Technical `impactAssessment` before acceptance. User-flow impact requires Product review and platform checks. Missing required platform checks leave the run unaccepted.
+- Inspect: `pipeline_status` reports current criterion/check status, open findings and reviewer provenance without binding sessions. `pipeline_reset` cancels without deleting history. `pipeline_prepare_merge` only prepares a report.
 - Read-only questions, review and existing test execution do not require a mutation pipeline. Scope-limited changes use the short pipeline; increased risk adds required checks.
 - MCP gates cover proxied calls. Native tool gates and continuation require observed host hooks. Plan Mode and user interruption take precedence.
 - Codegraph failures permit file-search fallback. The proxy refreshes the review graph before impact analysis; it must report failure rather than use stale results.
@@ -58,20 +59,9 @@ For Figma and Stitch tool names and authentication, read `docs/agents/design-con
 
 ---
 
-## 🧠 Memory Auto-Save Protocol (MANDATORY)
-
-Agent MUST auto-call `remember` in these situations — do NOT wait for user to ask:
-
-| Trigger | What to Save |
-|---|---|
-| After completing pipeline step `grill-me` | Design decisions + constraints discussed |
-| After completing pipeline step `diagnose` | Root cause + reproduction steps |
-| After completing pipeline step `tdd` | Implementation decisions + test patterns |
-| After completing pipeline step `improve-architecture` | Architecture improvements + quality findings |
-| User states a preference or convention | What they prefer + context |
-| Non-trivial bug resolved | Root cause, solution, affected files |
-
-**Pipeline checkpoint responses include `memory_action` field — follow it immediately.**
+## Memory
+
+Save durable decisions and reusable bug fixes when they will help future sessions. Do not write a memory record after every internal pipeline step or load the whole graph by default. A memory backend outage does not block unrelated core tools.
 
 ### Google Drive & Online Documents