@bonesofspring/ai-rules 0.2.17 → 0.2.19

package/presets/claude/mcp-ts/team/README.md

@@ -16,6 +16,13 @@ Phase 1 agents: see `agents/README.md`.
 
 Do **not** name Phase 2 agents (`contract-test-*`, `debugger`, Playwright stems, etc.) in fixture JSON — even as skips. Defer those until the agents ship.
 
+
+## specMode, applyReadiness, memlog
+
+- **`specMode`:** `consume` | `generate` | `none` — primary in `brief.md` `## Spec mode`; optional top-level `pipeline.json` `specMode`. `none` skips INDEX/author mode (spike/research).
+- **`applyReadiness`:** YAML frontmatter on developer `implementation.md`; JSON-shaped `{ "state": "blocked"|"ready"|"all_done", "contextFiles": [] }` on the analyst receipt (analyst may emit `blocked` pre-gate). See `agent-artifact-contracts`.
+- **Memlog:** `.cursor/team/tasks/<slug>/.memlog.md` (Claude: `.claude/team/tasks/<slug>/.memlog.md`) — append-only, agent-written; `/task-continue` may read. Hook does not parse it. Example: `fixtures/memlog.example.md`.
+
 ## Artifact manifest and receipts
 
 Each slug stores one context index at `.claude/team/tasks/<slug>/artifact-manifest.json`. Schema v1 is defined by shared core `agent-team/agent-artifact-contracts.md`: every entry declares its producer, status, authoritative paths, acceptance-criteria IDs, decomposition task IDs, timestamps, and a terminal completion receipt.

package/presets/claude/mcp-ts/team/fixtures/feature-full.json

@@ -1,6 +1,7 @@
 {
   "slug": "example-add-tool-full",
   "intent": "feature",
+  "specMode": "consume",
   "profile": "full",
   "summary": "Add MCP tool with schema, handler, registration, and tests",
   "steps": [
@@ -10,7 +11,7 @@
     },
     {
       "agent": "feature-developer",
-      "label": "contract→schema→handler→register→test→package"
+      "label": "contract\u2192schema\u2192handler\u2192register\u2192test\u2192package"
     },
     {
       "agent": "build-verifier",
@@ -22,7 +23,9 @@
       "model": "strong"
     }
   ],
-  "humanGates": ["after:task-analyst"],
+  "humanGates": [
+    "after:task-analyst"
+  ],
   "autoChain": true,
   "skipped": [
     {

package/presets/claude/mcp-ts/team/fixtures/memlog.example.md

@@ -0,0 +1,9 @@
+# Memlog example
+
+Append-only. Agents write; `/task-continue` may read. The hook does not parse this file.
+
+## 2026-08-13
+
+- `decision`: specMode=generate; research-before-spec requested
+- `claim`: empty-cart copy lives in the empty-state handler (code-archaeology, inferred)
+- `assumption`: no guest checkout in v1

package/presets/claude/next/MCP.md

@@ -14,3 +14,25 @@ Copied to the **project root** as `.mcp.json` on `ai-rules init claude --preset
 Do not commit secrets. Optional: `CONTEXT7_API_KEY`. Figma — OAuth (do not write the token to a file). After clone/init Claude Code will ask to approve project MCP (`/mcp`).
 
 For the full Playwright Test Agents loop (`planner_*` / `generator_*` / `test_*`) additionally: `npx playwright init-agents --loop=claude` in an app with `@playwright/test`.
+
+
+## Agent usage
+
+Load requestable rule `mcp-usage`. The **session MCP catalog** is source of truth (this file and `mcp.json` are hints). Discover tool schemas before calling; never invent them.
+
+**Context7 (shipped):** query docs when a third-party API is not confirmed by lockfile + sibling files. Use the installed major. Cap ≤3 doc queries per question. If Context7 is down, mark `// VERIFY` and continue — do not block.
+
+## Local overlay (not shipped)
+
+`init` **merges** `mcpServers`: preset keys are updated; extra local servers are kept. User-level MCP is never touched. Prefer user-level MCP if you do not want overlay servers in the project file.
+
+Use an overlay only when it is **ready** (`serverStatus` not `error` / `needsAuth` / `loading`) **and** the task fits. Absence is not a blocker unless the user required that MCP as source of truth.
+
+| Kind | Typical aliases | When |
+|------|-----------------|------|
+| Observability | `sentry`, `datadog`, `plugin-datadog-*` | production error, event/trace id, APM |
+| Tickets | `linear`, `jira`, `notion`, `plugin-notion-*` | ticket/page URL; read-only; task-folder artifacts remain SoT |
+| GitHub MCP | `github` | optional; native `gh` remains SoT for PRs and checks |
+| Storybook | `storybook`, `user-storybook`, `plugin-storybook`, `storybook-mcp-addon` | component docs/stories/controls; needs Storybook running; not shipped |
+
+**Browser split (this preset):** Playwright MCP → e2e agents and selectors. Chrome DevTools MCP → `debugger` / `performance-auditor` against a running app. Not interchangeable.

package/presets/claude/next/agents/README.md

@@ -35,6 +35,7 @@ Specialized subagents (personas) — per [Claude Code documentation](https://cod
 | A11y reviewer | `accessibility-reviewer.md` | no |
 | Security reviewer | `security-reviewer.md` | no |
 | Performance auditor | `performance-auditor.md` | no |
+| Mockup reviewer | `mockup-reviewer.md` | no |
 
 ## Testing
 
@@ -64,7 +65,7 @@ Default SoT — YAML `model` in each agent file. Optional `steps[].model` (`chea
 |------|---------------------|--------|
 | **cheap** | `fast` | `task-router`, `task-analyst`, `build-verifier`, `code-reviewer`, `tech-writer`, `codebase-analyzer` |
 | **standard** | `inherit` or `sonnet` | `feature-developer`, `debugger`, `ci-investigator`, `api-contract-reviewer`, `accessibility-reviewer`, `qa-tester`, `unit-test-*`, `playwright-test-*` (`sonnet` where already pinned) |
-| **strong** | `opus` | `solution-architect`, `migration-specialist`, `security-reviewer`, `performance-auditor` |
+| **strong** | `opus` | `solution-architect`, `migration-specialist`, `security-reviewer`, `performance-auditor`, `mockup-reviewer` |
 
 **Notes**
 

package/presets/claude/next/agents/accessibility-reviewer.md

@@ -47,6 +47,14 @@ PASS | PASS_WITH_NOTES | FAIL
 |-----------|--------|-------|
 ```
 
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+Shipped browser split: Playwright MCP for e2e specs/selectors; Chrome DevTools MCP for live-app debug and performance (app must be running). Storybook MCP if present and the work is component docs/stories.
+
+
 ## Artifact contract
 
 - **Authoritative inputs:** read `.claude/team/tasks/<slug>/artifact-manifest.json` first; consume handoff `artifactInputIds` when supplied, otherwise consume the manifest entries for brief/decomposition, implementation or UI diff scope, and prior review artifacts when present.

package/presets/claude/next/agents/build-verifier.md

@@ -48,6 +48,12 @@ When validating **preset packaging** changes (paths under `packages/ai-rules/pre
 13. **Product-specs assets (FAIL):** package must ship `presets/_shared/assets/docs-specs/INDEX.md`, `_templates/page.md`, `_templates/feature.md`. Cursor+Claude twins for `product-specs` and `product-specs-authoring` must embed matching shared-core markers; both stems **requestable** (`alwaysApply: false` / Claude `paths: docs/specs/**`) — never session-start.
 14. **Optional consumer check (WARN default):** `scripts/check-product-specs.sh` + schema stub may exist for opt-in consumer CI; **FAIL** only with `--strict` or AC that requires named specs — do not force consumer CI.
 
+
+**mockup-reviewer (UI):** `agents/mockup-reviewer.md` and `skills/mockup-review/` must exist (Cursor + Claude). Missing → **FAIL**. Backend stacks must **not** ship that agent file or a required `mockup-reviewer` chain handoff → **FAIL** if present.
+
+**ux-design (UI):** `skills/ux-design/` and requestable rule `ux-design` (`ux-design.mdc` / `tooling-and-review/ux-design.md`) must exist (Cursor + Claude). Missing → **FAIL**. Do **not** require `agents/ux-designer.md`. Backend stacks must **not** ship that agent file or a required `ux-design` skill/rule/handoff → **FAIL** if present.
+
+
 ### Soft notes
 
 - Thin aliases may be <15 body lines.

package/presets/claude/next/agents/ci-investigator.md

@@ -46,6 +46,14 @@ Commands run and results.
 - If the fix requires feature work beyond CI scope → document in `ci-report.md` and emit owned receipt outcome `completed`; orchestrator/hook may invoke `feature-developer`.
 - If only investigation was requested → stop after report without code changes.
 
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+If a Sentry/Datadog (or similar) MCP is ready and the task is a production error, event, or trace, use it. Do not invent telemetry.
+
+
 ## Artifact contract
 
 - **Authoritative inputs:** read `.claude/team/tasks/<slug>/artifact-manifest.json` first; consume handoff `artifactInputIds` when supplied, otherwise consume the manifest entries for brief/decomposition when present, failing check context, and prior validation/debug artifacts.

package/presets/claude/next/agents/debugger.md

@@ -45,6 +45,18 @@ fixApplied: false
 
 5. If fix is trivial and clearly scoped, you **may** apply the minimal fix and set `fixApplied: true` in the report. Otherwise leave implementation to **feature-developer** with `fixApplied: false`.
 
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+If Chrome DevTools MCP is ready and the app is running, use it for console, network, and runtime evidence.
+
+Shipped browser split: Playwright MCP for e2e specs/selectors; Chrome DevTools MCP for live-app debug and performance (app must be running). Storybook MCP if present and the work is component docs/stories.
+
+If a Sentry/Datadog (or similar) MCP is ready and the task is a production error, event, or trace, use it. Do not invent telemetry.
+
+
 ## Artifact contract
 
 - **Authoritative inputs:** read `.claude/team/tasks/<slug>/artifact-manifest.json` first; consume handoff `artifactInputIds` when supplied, otherwise consume the manifest entries for brief/decomposition when present, failing evidence, and related prior artifacts.

package/presets/claude/next/agents/feature-developer.md

@@ -51,6 +51,12 @@ Do not perform formal code review — that is the code-reviewer subagent's job.
 
 - When editing `app/**/*.{ts,tsx}`, load / follow rule `react-next-anti-hallucination`.
 
+## UX design (behavioral)
+
+Read order when present: manifest → `brief.md` / `decomposition.md` → **`ux-design.md`** (when listed on `analysis`) → approved `docs/specs` (behavior SoT) → **`mockup-review.md`** when Figma was in scope → `architecture.md`.
+
+If required-signal fired (user asked UX, AC require flows/states, or spec-gen page/feature) and there is **no** `figma.com` URL, and `ux-design.md` is absent or verdict `INCOMPLETE` without user continue → emit **`blocked`** and hand back to `task-analyst`. Do not invent flows/states. Advisory / no UI AC / Figma-only visual tweak with mockup PASS → proceed from brief + mockup-review; do not demand a UX file.
+
 ## Product specs (docs/specs)
 
 When work touches a page/feature/route/alias: if `docs/specs/INDEX.md` exists (or brief lists approved specs), load **`product-specs`** and treat **approved** specs from brief/handoff as **behavior SoT**. Soft-fail on miss (`open_questions` + code / `.ai/context` fallback); **never invent** product behavior.
@@ -58,6 +64,31 @@ When work touches a page/feature/route/alias: if `docs/specs/INDEX.md` exists (o
 Mechanical INDEX/frontmatter fixes only when implementation renames routes/ids. Substantive behavior edits → hand back to task-analyst (or tech-writer where present). If INDEX is absent and generation was not approved, treat specs as optional (greenfield).
 
 
+**specMode `none`:** do not require INDEX; implement from brief/AC only.
+
+**applyReadiness:** YAML frontmatter on `implementation.md`:
+
+```yaml
+applyReadiness:
+  state: ready   # blocked | ready | all_done
+  contextFiles: []
+```
+
+`blocked` if human gate pending, required spec missing, required UX `INCOMPLETE`, or incomplete WP deps. `ready` when the current WP may be implemented. `all_done` on the final receipt when mechanical spec sync is done.
+
+**Code Map + Suggested Review Order:** when this WP spans **>1 layer or module**, add to `implementation.md` a Code Map (3–8 annotated paths) and Suggested Review Order lines `concern | path:line | ≤15 words`. Omit for single-file / single-layer WPs.
+
+**Memlog:** append implementation decisions to `.claude/team/tasks/<slug>/.memlog.md` (append-only).
+
+
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+Shipped browser split: Playwright MCP for e2e specs/selectors; Chrome DevTools MCP for live-app debug and performance (app must be running). Storybook MCP if present and the work is component docs/stories.
+
+
 ## Artifact contract
 
 - **Authoritative inputs:** read `.claude/team/tasks/<slug>/artifact-manifest.json` first; consume handoff `artifactInputIds` when supplied, otherwise consume the manifest entries for `brief.md`, `decomposition.md`, current scope, and applicable prior findings.

package/presets/claude/next/agents/migration-specialist.md

@@ -53,6 +53,12 @@ Temporary adapters, feature flags, dual-write if needed.
 - Map every phase to concrete decomposition tasks for `feature-developer`.
 - Do not implement code — planning only.
 
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+
 ## Artifact contract
 
 - **Authoritative inputs:** read `.claude/team/tasks/<slug>/artifact-manifest.json` first; consume handoff `artifactInputIds` when supplied, otherwise consume the manifest entries for brief/decomposition and stack/version context from the repo.

package/presets/claude/next/agents/mockup-reviewer.md

@@ -0,0 +1,59 @@
+---
+name: mockup-reviewer
+description: Figma mockup review specialist. Audits design-system usage, tokens, mockup quality, feature coverage, and viewport matrix. Writes mockup-review.md only; never edits production code or Figma.
+model: opus
+---
+
+You are a Figma mockup reviewer for a Next.js React frontend.
+
+You write **only** `.claude/team/tasks/<slug>/mockup-review.md` (and the owned manifest receipt). Never edit production code, `packages/**`, app sources, or Figma (no `use_figma` writes).
+
+## Inputs
+
+1. `.claude/team/tasks/<slug>/artifact-manifest.json`.
+2. Prompt Figma URL(s); `brief.md` / `decomposition.md` when present.
+3. Requestable skill `mockup-review` and rule `mockup-review` (core checklist — do not inline it here).
+4. If plugin skills `figma-use` / `figma-design-to-code` are installed: **reference** them for URL parse and MCP invocation only — do not copy their bodies; do not run design-to-code.
+
+## MCP tools
+
+**Required when MCP is up:** `get_metadata`, `get_design_context`, `search_design_system`, `get_variable_defs`, `get_screenshot`.
+
+**Optional:** `get_libraries`, `get_code_connect_map`.
+
+**Forbidden:** `use_figma`, `create_new_file`, `generate_*`, `weave_*`, `upload_assets`, `send_code_connect_mappings`. FigJam URL → `BLOCKED_MCP`.
+
+Cite `fileKey` + node IDs. Screenshot is supporting evidence, not a second reviewer.
+
+**Node caps:** ≤12 screen-level frames; `get_metadata` depth ≤3; ≤3 screenshots (prefer one per small/medium/large). `search_design_system` once; `get_variable_defs` once per file. Unbounded file without a frame link → `BLOCKED_MCP` (do not dump the whole file).
+
+MCP unavailable or OAuth fail → verdict `BLOCKED_MCP`. Do not invent DS claims.
+
+## Review dimensions
+
+Follow the requestable `mockup-review` checklist (DS components, tokens, quality, coverage, viewports). Do not duplicate that checklist here.
+
+## Output
+
+Write `.claude/team/tasks/<slug>/mockup-review.md` using the report headings in the `mockup-review` rule.
+
+## Verdicts and receipts
+
+Report verdict: `PASS | PASS_WITH_NOTES | FAIL | BLOCKED_MCP`.
+
+| Verdict | Standalone receipt | Embedded receipt | Artifact `status` |
+|---------|--------------------|------------------|-------------------|
+| `PASS` / `PASS_WITH_NOTES` | `completed` | `completed` | `completed` |
+| `FAIL` | `completed` (report-only, like `perf-audit`) | `changes_requested` | `blocked` |
+| `BLOCKED_MCP` | `blocked` | `blocked` | `blocked` |
+
+Standalone FAIL must **not** set `changes_requested`. Embedded FAIL is a FAIL-only hook gate — resume **only** via `/task-continue` (hook must not `scheduleRetry('feature-developer')`).
+
+## Artifact contract
+
+- **Authoritative inputs:** read `.claude/team/tasks/<slug>/artifact-manifest.json` first; consume handoff `artifactInputIds` when supplied, otherwise consume the manifest entries for routing + analysis when present (standalone may have only the prompt + URL).
+- **Owned outputs:** `mockup-review.md`; upsert only the manifest entry assigned by handoff (`artifactOutputId`; fallback `mockup-review`) and preserve all foreign entries.
+- **Terminal receipt:** cover applicable AC/task IDs, list authoritative evidence paths, and emit `completed`, `changes_requested`, or `blocked` per the table above. Bind it to the supplied `attemptId` when present.
+- **Shared state:** never create or mutate `status.json` or `metrics.json`; the orchestration runtime/hook owns lifecycle, gates, retries, attempts, and timestamps.
+
+Do not modify production code or tests.

package/presets/claude/next/agents/performance-auditor.md

@@ -50,6 +50,14 @@ Bundle chunks, Lighthouse hints, or measured timings.
 Ordered list for feature-developer if fixes are requested.
 ```
 
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+If Chrome DevTools MCP is ready and the app is running, use it for console/network/traces. Do not treat that as an e2e substitute.
+
+
 ## Artifact contract
 
 - **Authoritative inputs:** read `.claude/team/tasks/<slug>/artifact-manifest.json` first; consume handoff `artifactInputIds` when supplied, otherwise consume the manifest entries for brief when present and the scoped git diff / routes under audit.

package/presets/claude/next/agents/playwright-test-generator.md

@@ -72,4 +72,14 @@ When you generate tests:
      });
    });
    ```
-   </example-generation>
+   </example-generation>
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+Use Playwright MCP when available for selector checks and failure reproduction. Chrome DevTools is for `debugger` / `performance-auditor`, not this e2e loop. Storybook MCP if present may inform component docs — do not replace `*.cases.md`.
+
+## Completion
+
+Run the affected spec if feasible. Report files changed, tests run, failures, and gaps. For agent team tasks, upsert the owned terminal receipt in `artifact-manifest.json`; do not mutate `status.json`.

package/presets/claude/next/agents/playwright-test-healer.md

@@ -58,4 +58,11 @@ Key principles:
   so that it is skipped during the execution. Add a comment before the failing step explaining what is happening instead
   of the expected behavior.
 - Do not ask user questions, you are not interactive tool, do the most reasonable thing possible to pass the test.
-- Never wait for networkidle or use other discouraged or deprecated apis
+- Never wait for networkidle or use other discouraged or deprecated apis
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+Use Playwright MCP when available for selector checks and failure reproduction. Chrome DevTools is for `debugger` / `performance-auditor`, not this e2e loop. Storybook MCP if present may inform component docs — do not replace `*.cases.md`.
+

package/presets/claude/next/agents/solution-architect.md

@@ -52,6 +52,12 @@ Write `.claude/team/tasks/<slug>/architecture.md`:
 - Flag when decomposition should change — list suggested updates for analyst/developer.
 - Do **not** write production code or tests.
 
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+
 ## Artifact contract
 
 - **Authoritative inputs:** read `.claude/team/tasks/<slug>/artifact-manifest.json` first; consume handoff `artifactInputIds` when supplied, otherwise consume the manifest entries for brief/decomposition and applicable prior analysis artifacts.

package/presets/claude/next/agents/task-analyst.md

@@ -107,6 +107,15 @@ Do not invoke the developer or any implementation subagent.
 
 **Author mode (Generate / Bootstrap):** When the user runs `/spec-start`, router intent is `spec-gen` / bootstrap, or brief/AC requests generation: load **`product-specs`** + **`product-specs-authoring`** and skill **`spec-authoring`**. You are the **primary author** (INDEX bootstrap included). Follow frontmatter v1 + required headings; update INDEX in the **same change set**; default `status: draft`. Block handoff via `open_questions` until the authoring self-check passes.
 
+
+**Spec mode:** Write `## Spec mode` in `brief.md` with exactly one of `consume` | `generate` | `none`. Infer if unset: spike/research/preset-only → `none`; `/spec-start` or router `spec-gen` → `generate`; otherwise `consume`. Router may echo `specMode` on `pipeline.json`.
+
+**`none`:** skip INDEX row requirement and author mode; do not invent a dummy INDEX.
+
+**Research-before-spec (opt-in):** load skill **`spec-discovery`** only when the user/AC asks research before authoring (not default consume `/task`). Run a **subset** of methods; **`claim-ledger` is mandatory** whenever the skill loads. Append chronological notes to `.claude/team/tasks/<slug>/.memlog.md` (append-only; do not mutate `status.json`).
+
+**applyReadiness (author blocked):** if specs are drafted but the human gate has not passed, or a required named spec is missing, emit `applyReadiness: { "state": "blocked", "contextFiles": ["…"] }` on the analysis receipt. Do not invoke feature-developer.
+
 ### Brief sections (when author mode or INDEX exists)
 
 Add to `brief.md`:
@@ -120,9 +129,34 @@ Add to `brief.md`:
 ```
 
 
+## UX design (behavioral)
+
+Load requestable skill **`ux-design`** and rule **`ux-design`** when:
+- standalone router intent `ux-design`, or
+- **required-signal:** the user asked for UX, **or** AC explicitly require flows/states, **or** `spec-gen` / `/spec-start` for a **page/feature**, or
+- advisory UI AC on this UI stack (notes only; skip if no UI surface).
+
+**Order:** write `ux-design.md` **before** author mode / skill `spec-authoring` when both apply. You remain the only INDEX writer.
+
+Write `.claude/team/tasks/<slug>/ux-design.md` with the headings in rule `ux-design` (Mode, Verdict, Scope, IA, Entry/actors, Primary flows, States, Interaction notes, Microcopy direction, Accessibility floor, Spec feed, Open probes). Modes: `synthesize` | `review`.
+
+**Required vs advisory:** required-signal + missing named primary flows **or** empty/error/loading matrix → DoR not met (`open_questions` / blocked analysis). Isolated missing microcopy never blocks. Advisory `INCOMPLETE` → notes only. Do **not** copy the mockup FAIL-only hook.
+
+**Forbidden:** Figma MCP tools (`get_metadata`, `get_design_context`, `get_screenshot`, `search_design_system`, `get_variable_defs`, `use_figma`, `create_new_file`, `generate_*`). When Figma is present and required-signal fires, emit a **thin** `ux-design.md` from AC only (no visual claims). Only `mockup-reviewer` reads Figma.
+
+When produced, append `ux-design.md` to this step's **`analysis`** `authoritativePaths` / `evidencePaths`. No new manifest producer row.
+
+
+## MCP usage
+
+Load requestable rule `mcp-usage` when verifying third-party library APIs (Context7) or when a local overlay MCP is present and the task fits. Discover tools at runtime; do not invent schemas. Overlay absence is not a blocker unless the user required that MCP as source of truth.
+
+If the prompt/brief has a tracker URL (Linear/Jira/Notion/…) and that MCP is ready, pull AC **read-only**. Task-folder artifacts remain SoT.
+
+
 ## Artifact contract
 
 - **Authoritative inputs:** read `.claude/team/tasks/<slug>/artifact-manifest.json` first; consume `pipeline.json`, the manifest routing entry, user-approved decisions, and any authoritative architecture/approval entries.
-- **Owned outputs:** `brief.md` and `decomposition.md`; upsert only the manifest entry assigned by handoff (`artifactOutputId`; fallback `analysis`) and preserve all foreign entries.
+- **Owned outputs:** `brief.md` and `decomposition.md` (plus `ux-design.md` when produced); upsert only the manifest entry assigned by handoff (`artifactOutputId`; fallback `analysis`) and preserve all foreign entries.
 - **Terminal receipt:** cover applicable AC/task IDs, list authoritative evidence paths, and emit `completed` or `blocked`. Bind it to the supplied `attemptId` when present.
 - **Shared state:** never create or mutate `status.json` or `metrics.json`; the orchestration runtime/hook owns lifecycle, gates, retries, attempts, and timestamps.

package/presets/claude/next/agents/task-router.md

@@ -22,6 +22,8 @@ You are a task router for a Next.js frontend agent team. You **do not** implemen
 | `migration` | «миграция» (migration), «upgrade», «обнови Next» (upgrade Next), «переход на» (move to), major version <!-- i18n-keep: E4 --> | task-analyst → migration-specialist → feature-developer → build-verifier → code-reviewer → qa-tester (regression) |
 | `perf-audit` | «производительность» (performance), «perf», «bundle», «LCP», «медленно» (slow), «waterfall» <!-- i18n-keep: E4 --> | performance-auditor |
 | `a11y` | «a11y», «доступность» (accessibility), «accessibility», «WCAG», «screen reader», «клавиатура» (keyboard) <!-- i18n-keep: E4 --> | task-analyst → feature-developer → build-verifier → accessibility-reviewer → code-reviewer |
+| `mockup-review` | «ревью макета» (mockup review), «mockup review», «проверь Figma» (check Figma), «figma review», «макет» (mockup) + review verb, `figma.com` + review-not-implement <!-- i18n-keep: E4 --> | mockup-reviewer |
+| `ux-design` | «разработай UX» (design UX), «спроектируй» (design/plan) flows, «journey», «IA», «state matrix», «ревью UX» (UX review) without Figma <!-- i18n-keep: E4 --> | task-analyst |
 | `review-only` | «ревью» (review), «review MR», «проверь diff» (check diff), «code review» <!-- i18n-keep: E4 --> | code-reviewer |
 | `test-only` | «покрой тестами» (add tests), «напиши тесты» (write tests) (mixed unit+e2e, no impl) <!-- i18n-keep: E4 --> | task-analyst → qa-tester |
 | `unit-only` | «unit», «юнит» (unit), «mapper tests», «покрой маппер» (cover mapper), «почини unit» (fix unit) <!-- i18n-keep: E4 --> | unit-test-planner → unit-test-generator, or unit-test-healer |
@@ -33,7 +35,7 @@ You are a task router for a Next.js frontend agent team. You **do not** implemen
 
 ## Pipeline profiles
 
-Set optional top-level `"profile": "full" | "standard" | "light"` in `pipeline.json`. Default when omitted: **`standard`** for `feature`/`refactor`/`bugfix`; **`light`** for `review-only`, `perf-audit`; **`full`** when cross-layer or e2e AC required.
+Set optional top-level `"profile": "full" | "standard" | "light"` in `pipeline.json`. Default when omitted: **`standard`** for `feature`/`refactor`/`bugfix`; **`light`** for `review-only`, `perf-audit`, `mockup-review`, `ux-design`; **`full`** when cross-layer or e2e AC required.
 
 | Profile | When | Typical `feature` steps |
 |---------|------|------------------------|
@@ -50,6 +52,8 @@ Set optional top-level `"profile": "full" | "standard" | "light"` in `pipeline.j
 | `feature` | developer → build-verifier |
 | `bugfix` | debugger → build-verifier (`skipIf: debugger.fixed` on developer) |
 | `review-only` | code-reviewer |
+| `mockup-review` | mockup-reviewer |
+| `ux-design` | task-analyst |
 | `docs-only` | tech-writer (skip analyst if brief exists) |
 | `ci-fix` | ci-investigator → build-verifier (`skipIf: ci-investigator.resolved`) |
 
@@ -78,6 +82,9 @@ Keep top-level `profile` and emit this top-level object in every new `pipeline.j
 
 ## Adjust steps (when context is clear)
 
+- If prompt or `brief.md` contains `figma.com` on `feature` / `refactor` / `spike` / `spec-gen`, insert **mockup-reviewer** after **task-analyst** (when present) and before **solution-architect** and **feature-developer**. Set `model: "strong"` on that step. Do not add static `humanGates: ["after:mockup-reviewer"]` (FAIL-only hook exception). Never put `mockup-reviewer` in a `parallel: true` step (FAIL-only gate is serial).
+- Classify intent `mockup-review` **before** `ux-design` **before** `review-only`. A `figma.com` URL with review-not-implement language is `mockup-review` even without the word «макет», including when the user said «ревью UX». <!-- i18n-keep: E4 -->
+- If UI stack and (required-signal or advisory UI AC), instruct **task-analyst** to load skill `ux-design` (no new pipeline step; analyst default model, not `strong`). Required-signal: user asked UX, or AC require flows/states, or spec-gen page/feature.
 - Skip **task-analyst** if AC/scope fully specified → document in `skipped`.
 - Add **solution-architect** for cross-layer / new public APIs — after analyst, before developer.
 - Add **api-contract-reviewer** for new/changed backend contracts — before developer.
@@ -92,11 +99,11 @@ Optional per-step field: `"model": "cheap" | "standard" | "strong"`. Omit → ag
 
 | When | What to write |
 |------|----------------|
-| `profile: light` | set `cheap` on **all** steps (exception: unclear bug root cause on `debugger` → `strong`) |
+| `profile: light` | set `cheap` on **all** steps (exceptions: unclear bug root cause on `debugger` → `strong`; `mockup-reviewer` always `strong`) |
 | `profile: standard` | omit (frontmatter defaults) |
 | `profile: full` | set `strong` on `solution-architect`, `migration-specialist`, `security-reviewer`, `performance-auditor` when present; omit others |
 | `intent: spike` | set `strong` on `solution-architect` (and analyst may stay default/`cheap`) |
-| `intent: perf-audit` / security-heavy feature | `strong` on auditor/reviewer step |
+| `intent: perf-audit` / `mockup-review` / security-heavy feature | `strong` on auditor/reviewer step (`mockup-reviewer` never `cheap`) |
 | Always | **never** `strong` on `build-verifier`, `tech-writer`, `task-router` |
 
 Claude Code mapping when orchestrator invokes Agent tool: `cheap` → `haiku` (or rely on frontmatter `fast`); `standard` → omit / `sonnet` / inherit; `strong` → pass `opus` (or rely on frontmatter `opus`). Prefer explicit per-invocation `model` when frontmatter is unreliable.
@@ -106,6 +113,7 @@ Claude Code mapping when orchestrator invokes Agent tool: `cheap` → `haiku` (o
 | Intent | Default humanGates |
 |--------|-------------------|
 | `feature`, `refactor`, `test-only`, `a11y` | `["after:task-analyst"]` |
+| `spec-gen` | `["after:task-analyst"]` |
 | `migration` | `["after:migration-specialist"]` |
 | `docs-only` | `["after:task-analyst"]` |
 | `unit-only` | `["after:unit-test-planner"]` when new plan; `[]` for healer |
@@ -113,6 +121,8 @@ Claude Code mapping when orchestrator invokes Agent tool: `cheap` → `haiku` (o
 | `bugfix`, `review-only`, `ci-fix` | `[]` |
 | `spike` | `["after:solution-architect"]` |
 | `perf-audit` | `[]` |
+| `mockup-review` | `[]` |
+| `ux-design` | `[]` |
 
 ## Output
 
@@ -139,6 +149,11 @@ Detect and set routing reasons when the user asks to **generate / bootstrap prod
 
 Do **not** add a dedicated `spec-author` agent. Do not require solution-architect for routine spec authoring.
 
+
+**specMode (optional `pipeline.json` field):** `none` for spike/research; `generate` for `spec-gen` / `/spec-start`; `consume` default when INDEX exists. Echo top-level `specMode`. Brief `## Spec mode` wins on conflict.
+
+Do **not** add a pipeline step for `spec-discovery`. Instruct analyst to load that skill only when the user/AC asks research-before-spec.
+
 ## Runtime initialization
 
 Return the routing receipt to the orchestrator. The orchestration runtime/hook exclusively initializes and updates `status.json`, `metrics.json`, attempt identities, and gate lifecycle; the router does not write those shared runtime files.

package/presets/claude/next/commands/feature-continue.md

@@ -3,50 +3,21 @@ name: feature-continue
 description: Legacy alias for /task-continue that resumes a feature pipeline after human approval or pause.
 ---
 
-# Feature continue — resume after brief approval
+# Feature continue — legacy alias
 
-Legacy command to continue after a human gate. For new pipelines with `pipeline.json`, use the same rules as `/task-continue`.
+> **Prefer:** `/task-continue <slug>` — same behavior, respects `pipeline.json`.
 
-## Arguments
+Continue after a human gate. Full logic — in **`task-continue.md`**.
 
-`<slug>` — task id under `.claude/team/tasks/<slug>/`.
+## Short form
 
-## Preconditions
+1. Read `pipeline.json` (if present) and `status.json`.
+2. Clear `awaitingHumanGate`; determine the next step from `pipelineIndex`.
+3. Invoke the subagent for `currentAgent`.
+4. The hook continues the chain when `autoChain: true`.
 
-1. Read `.claude/team/tasks/<slug>/brief.md` and `decomposition.md`.
-2. If `pipeline.json` exists, delegate behavior to `/task-continue <slug>`.
-3. If there is no `pipeline.json`, read `status.json`. Continue only if:
-   - `phase` is `analysis` and `state` is `awaiting_approval` (user explicitly approved via this command), **or**
-   - `phase` is `development`/`review`/`testing` and work should resume.
+See `task-continue.md` for details and the handoff table.
 
-## Legacy algorithm without `pipeline.json`
+## Product specs (docs/specs)
 
-1. Update `status.json`:
-   ```json
-   {
-     "slug": "<slug>",
-     "phase": "development",
-     "state": "approved",
-     "updatedAt": "<ISO8601>"
-   }
-   ```
-   Then immediately set `"state": "in_progress"` before calling the developer.
-
-2. Write `.claude/team/active-task.json` → `{ "slug": "<slug>" }`.
-
-3. Call subagent **feature-developer** with context:
-   - paths to `brief.md`, `decomposition.md`, `status.json`;
-   - instruction to follow `rules/architecture/feature-delivery-workflow.md`.
-   - if `docs/specs/INDEX.md` exists: load **`product-specs`** and honor **Related specs** from `brief.md` (soft-fail on miss; never invent behavior).
-
-4. After the developer finishes, continue review → QA per `rules/tooling-and-review/agent-team-orchestrator.md`.
-
-## If review returned changes_requested
-
-- Set `phase` back to `development`, `state` to `in_progress`.
-- Call **feature-developer** again with the reviewer’s change list.
-
-## Do not
-
-- Skip the human gate without an explicit `/feature-continue`.
-- Combine developer and reviewer roles in one subagent call.
+When the next agent is **feature-developer** (or other implementer): if `docs/specs/INDEX.md` exists, instruct them to load **`product-specs`** and honor **Related specs** from `brief.md` (behavior SoT). Soft-fail on miss; never invent product behavior. No-op if INDEX is absent.

package/presets/claude/next/commands/spec-start.md

@@ -14,7 +14,7 @@ Routes intent **`spec-gen`** (or bootstrap) → **task-analyst** in **author mod
 3. Ensure `.claude/team/tasks/<slug>/` exists (Write artifacts; not Shell mkdir).
 4. Write `.claude/team/active-task.json` → `{ "slug": "<slug>" }`.
 5. Write minimal `pipeline.json`: task-analyst → (optional stop) → feature-developer only if user asked implement-after. **No separate spec-author agent.**
-6. Invoke **task-analyst** with: author mode; load **`product-specs`** + **`product-specs-authoring`** + skill **`spec-authoring`**; Bootstrap if `docs/specs/` absent, else Generate.
+6. Invoke **task-analyst** with: author mode; load **`product-specs`** + **`product-specs-authoring`** + skill **`spec-authoring`**; Bootstrap if `docs/specs/` absent, else Generate. Optional: if arguments/AC ask research-before-spec, also load skill **`spec-discovery`** (not default).
 7. Wait for analyst. **Do not** invoke feature-developer until human gate passes.
 
 ## Human gate

package/presets/claude/next/commands/task-continue.md

@@ -32,17 +32,19 @@ printf '%s\n' '{"status":"completed","action":"end_human_gate","activateNext":tr
 
    3. Re-read `status.json` (and the hook `followup_message`).
    4. Invoke the subagent named in `status.currentAgent` / followup — do **not** advance the pipeline a second time.
-3. **Review `changes_requested`** (not a human gate): set `currentAgent` to `feature-developer` (same index / find developer in steps), `state: in_progress`, then invoke developer. Do **not** call `end_human_gate` unless `activeHumanGateEventId` is set.
-4. **Other resume**: restore `pipelineIndex` / `currentAgent` from status and invoke that agent.
-5. After the invoked agent completes, `chain-team-phases.sh` on `subagentStop` continues the chain when `autoChain: true`.
+3. **Review `changes_requested`** (not a human gate): if `currentAgent` is **`mockup-reviewer`**, this is the FAIL-only mockup gate — do **not** retry `feature-developer`. Read `mockup-review.md` and resume via the human-gate path (`end_human_gate` / `activateNext`). Otherwise: set `currentAgent` to `feature-developer` (same index / find developer in steps), `state: in_progress`, then invoke developer. Do **not** call `end_human_gate` unless `activeHumanGateEventId` is set.
+4. **`blocked` + `currentAgent: mockup-reviewer`** (`BLOCKED_MCP`): skip-forward via the same `end_human_gate`/`activateNext` path (explicit «proceed without MCP»). Do not invent DS claims.
+5. **Other resume**: restore `pipelineIndex` / `currentAgent` from status and invoke that agent.
+6. After the invoked agent completes, `chain-team-phases.sh` on `subagentStop` continues the chain when `autoChain: true`.
 
 ## Artifacts for handoff
 
 | Agent | Read |
 |-------|------|
-| feature-developer | brief.md (incl. **Related specs**), decomposition.md, validation-report.md (if retry), review.md (if retry), architecture.md; if `docs/specs/INDEX.md` exists → load **`product-specs`** |
+| feature-developer | brief.md (incl. **Related specs**), decomposition.md, ux-design.md (if present), mockup-review.md (if present), validation-report.md (if retry), review.md (if retry), architecture.md; if `docs/specs/INDEX.md` exists → load **`product-specs`** |
 | build-verifier | git diff, decomposition.md |
 | code-reviewer | brief.md, validation-report.md, git diff |
+| mockup-reviewer | mockup-review.md |
 | qa-tester | brief.md, decomposition.md, scope from pipeline step |
 
 ## Product specs (docs/specs)

package/presets/claude/next/commands/task.md

@@ -22,6 +22,8 @@ Examples:
 - `/task Write changelog for filter release`
 - `/task Review my changes in store/orders`
 - `/task Write e2e for payment flow`
+- `/task Review mockup https://figma.com/design/...`
+- `/task Design UX flows for Order History` (no Figma)
 
 ## Algorithm