@bonesofspring/ai-rules 0.2.21 → 0.2.23

package/presets/mavis/next/rules/types-jsdoc.md

@@ -0,0 +1,43 @@
+# Type documentation in `app/src/types`
+
+When adding or substantially changing types in this layer, **follow the JSDoc style already adopted in the repo** (see examples: `User.types.ts`, `Common.types.ts`, `ServerValidation.types.ts`, `TreatmentPlan/*.types.ts`).
+
+## Language and form
+
+- Comment text — **in Russian** (repo locale convention for domain/API field descriptions), brief and to the point: what the type or field means in the domain or API contract.
+- **Do not rely** on tags like `@param`, `@returns`, `@see`, `@deprecated` for type descriptions — they are **not used** in this layer; plain text in `/** … */` is enough.
+
+## Exported `type` / `interface`
+
+- Immediately **before the declaration** — a `/** … */` block.
+- If you need to explain data source, constraints, or non-obvious details — a **second paragraph** in the same block (in JSDoc, separate paragraphs with a blank line between text lines).
+- For **nested** objects in the same file, document each type with its own block above its declaration.
+
+## Fields
+
+- For **every** public property — a **single-line** `/** … */` on the line directly **above** the field.
+- If a field **does not come from the backend as-is** but is computed or enriched on the frontend (mapper, selector, UI) — start the description with the **`[computed]`** prefix (as in `TUserProfile`, treatment plan types).
+- For data formats, state this **in the text**: e.g. date `YYYY-MM-DD`, a display string example in quotes.
+
+Example field comment (Russian locale — keep Cyrillic in type docs): <!-- i18n-keep: E5 -->
+
+```ts
+/** Дата рождения в формате YYYY-MM-DD */ <!-- i18n-keep: E5 -->
+birthDate: string
+```
+
+## Classes and other declarations
+
+- For **classes** (e.g. transport errors), when changing public API — a brief block above the class and **comments on public fields** using the same rules as interface properties.
+
+## `enum`
+
+- In `enums.ts`, members are often **without JSDoc** historically; for new enums a short description **above the enum** is acceptable if purpose is not obvious from the name. UI labels — as established, via `*Names` objects next to the enum.
+
+## Agent practice
+
+- When adding a new type or field, **do not leave** new public fields unexplained if meaning is not 100% clear from the name.
+- When editing a file that already has such comments, **maintain the same style**, not mixed with English-only or tag-heavy blocks.
+- Do not bloat comments: one or two sentences per type, one line per field is normal; exception — a genuinely complex domain caveat.
+
+See also imports and barrel: `public-imports.md`.

package/presets/mavis/next/rules/ux-design.md

@@ -0,0 +1,99 @@
+<!-- shared-core: review/ux-design.md -->
+
+# UX design checklist
+
+On-demand behavioral UX checklist (IA, named flows, state matrix). **Procedure:** skill `ux-design`. **Producer:** `task-analyst` (not a pipeline agent). Do not inline this checklist into the analyst prompt.
+
+Stack-agnostic: no app-source globs, Playwright, or `xcodebuild`. Form-factor **examples** live in adapter **Stack notes** only.
+
+## When to load
+
+- Standalone `/task` intent `ux-design` (modes `synthesize` | `review`; profile `light`; steps only `task-analyst`)
+- Embedded on a UI-capable stack when **required-signal** or advisory UI AC fires on `feature` / `refactor` / `spike` / `spec-gen`
+- Required-signal (any one): the user asked for UX, **or** AC explicitly require flows/states, **or** intent is `spec-gen` / `/spec-start` for a **page/feature**. Else **advisory** (notes only; skip if no UI surface)
+- **Do not** load for backend-only AC even on a UI stack. Classify `mockup-review` (Figma) before `ux-design` before `review-only`
+
+## Inputs
+
+1. Task `artifact-manifest.json`, then prompt / `brief.md` / `decomposition.md` when present
+2. Skill `ux-design` + this rule
+3. Approved `docs/specs` only as product-behavior SoT **after** UX synthesis maps via Spec feed — do not write INDEX here
+
+## Forbidden Figma MCP
+
+Do **not** call Figma MCP. Only `mockup-reviewer` reads Figma. Forbidden by name: `get_metadata`, `get_design_context`, `get_screenshot`, `search_design_system`, `get_variable_defs`, `use_figma`, `create_new_file`, `generate_*`.
+
+When `figma.com` is present and required-signal fires, emit a **thin** `ux-design.md` from AC only (IA/flows the mockup must cover; no visual or DS claims). If the user only pasted a URL and AC are visual, the UX file may be omitted; `mockup-review.md` coverage stands.
+
+Do **not** write BMAD-style `DESIGN.md` or a second visual spine. One experience artifact: `ux-design.md`.
+
+## Modes
+
+| Mode | When |
+|------|------|
+| `synthesize` | Design UX from a brief/AC (standalone «разработай UX» / embedded required-signal) | <!-- i18n-keep: E4 -->
+| `review` | Critique a brief/wireframe/description **without** Figma («ревью UX»). Figma + review-not-implement remains intent `mockup-review` | <!-- i18n-keep: E4 -->
+
+## Required vs advisory
+
+- **Required** when required-signal fires (user-asked-ux / ac-requires-flows-states / spec-gen-page-feature)
+- **Advisory** otherwise (UI AC without those signals). Advisory `INCOMPLETE` → notes only; DoR may complete
+- Skip the skill when there is no UI surface or the prompt is empty of UX-relevant AC
+
+## Verdicts
+
+`PASS` | `PASS_WITH_NOTES` | `INCOMPLETE`. Never copy the mockup FAIL-only hook (`changes_requested` + do-not-retry-developer). Incomplete UX is an analyst/DoR problem.
+
+| Verdict | Required step | Advisory step | Standalone intent `ux-design` |
+|---------|---------------|---------------|-------------------------------|
+| `PASS` / `PASS_WITH_NOTES` | Analysis may complete DoR | Analysis may complete | `completed` |
+| `INCOMPLETE` | DoR **not** met: `open_questions` / blocked analysis; existing `after:task-analyst` gate | Notes only; DoR may complete | `completed` (report-only) |
+
+**Block (`INCOMPLETE` as DoR failure) only when the step was required and primary flows *or* the empty/error/loading matrix are missing.** Isolated missing microcopy or optional protagonist names never block.
+
+## Surface-closure
+
+Each stated need → a surface; each surface → a flow that lands there. If closure fails, list probes under **Open probes** — **do not invent** screens, flows, or copy to fill gaps.
+
+## Spec feed
+
+Analyst remains the **only** writer of `docs/specs` INDEX and page/feature files. Map `ux-design.md` → templates **after** UX synthesis (UX before `spec-authoring` author mode):
+
+| `ux-design.md` heading | Page template | Feature template |
+|------------------------|---------------|------------------|
+| Scope + IA | **Overview** | **Overview** |
+| Entry / actors | **Entry / actors** | — (page-level) |
+| Primary flows | **Primary flows** | **Shared behavior / constraints** when a flow is cross-page |
+| States | **States** (loading/empty/error/success) | **Cross-cutting states** when shared |
+| Interaction notes + microcopy direction | Inform **Acceptance criteria**; leftover → **Edge cases** | Shared constraints |
+| Accessibility floor (behavioral) | Testable AC without pixels; else Edge cases | Shared constraints |
+| Surface-closure probes | Analyst `open_questions` — **do not** invent spec rows | Same |
+
+## Manifest
+
+When produced, append `ux-design.md` to the **`analysis`** entry `authoritativePaths` and `evidencePaths`. Do **not** add a new default producer row for v1 (`ux-designer` is deferred).
+
+## Report headings
+
+Write `ux-design.md` with these headings (must match the architecture artifact contract):
+
+```markdown
+# UX design: <slug>
+## Mode
+synthesize | review
+## Verdict
+PASS | PASS_WITH_NOTES | INCOMPLETE
+## Scope
+- Form factor: web | ios | android | mixed
+- Figma present: yes | no
+- Required-signal: user-asked-ux | ac-requires-flows-states | spec-gen-page-feature | advisory
+## Information architecture
+## Entry / actors
+## Primary flows
+## States
+## Interaction notes
+## Microcopy direction
+## Accessibility floor (behavioral)
+## Spec feed
+## Open probes
+```

package/presets/mavis/next/skills/README.md

@@ -0,0 +1,21 @@
+# `.mavis/skills` (preset next)
+
+Project skills for Cursor. They keep long workflows out of always-on rules while preserving the same team conventions as the Claude preset.
+
+| Skill | When to use |
+|-------|-------------|
+| `feature-delivery` | End-to-end feature work across types, API, store, UI, tests |
+| `spec-authoring` | Author or bootstrap docs/specs (INDEX, page/feature templates); `/spec-start` / analyst author mode |
+| `spec-discovery` | Opt-in research-before-spec in analyst author mode (claim ledger; not default `/task` consume) |
+| `mockup-review` | Figma mockup review: DS, coverage, viewports (standalone or embedded) |
+| `ux-design` | Behavioral UX synthesis/review via task-analyst (standalone or embedded; no Figma MCP) |
+| `code-review` | MR/diff review using project architecture and test standards |
+| `debug-investigation` | Reproducing and fixing bugs with root-cause evidence |
+| `ci-investigation` | Diagnosing and fixing failing CI checks on PRs/branches |
+| `unit-testing` | Planning, generating, or healing unit test coverage |
+| `playwright-e2e` | Planning, generating, or healing Playwright e2e coverage |
+| `technical-retro` | Technical retrospective for a task, sprint, incident, or agent pipeline |
+| `write-adr` | Author an Architecture Decision Record (context / decision / consequences) |
+| `react-performance` | Implementer checklist: bundle, RSC/client, memo, waterfalls, CWV (not a perf audit agent) |
+
+Skills are copied to `.mavis/skills/` by `ai-rules init mavis --preset next`.

package/presets/mavis/next/skills/ci-investigation/SKILL.md

@@ -0,0 +1,39 @@
+# CI Investigation
+
+## Workflow
+
+1. Identify the **first failing check** (lint, type-check, unit, e2e, build).
+2. Capture error output — ignore cascading failures until root cause is found.
+3. Reproduce locally with the same commands CI uses from `app/`:
+   - `lint:js`, `lint:css`, `type-check`
+   - targeted unit test file or suite
+   - affected Playwright spec if e2e failed
+4. Inspect git diff for changed files related to the failure.
+5. Apply **minimal fix** — no unrelated refactors.
+
+## Artifact
+
+Write `ci-report.md` under the task slug folder:
+
+- Failed check name and link/log excerpt
+- Root cause classification
+- Fix applied (or handoff to feature-developer)
+- Validation commands and results
+
+
+## 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.
+
+
+## Handoff Rules
+
+- Config-only or one-line fixes → fix in place, re-run check.
+- Missing tests or architecture work → document and route to `feature-developer` or `qa-tester`.
+- Flaky e2e → prefer `playwright-test-healer` after ci-investigator identifies the spec.
+
+## Validation
+
+Always re-run the failing command locally before marking complete. If environment blocks execution, state the exact command and blocker.

package/presets/mavis/next/skills/code-review/SKILL.md

@@ -0,0 +1,39 @@
+# Code Review
+
+## Inputs
+
+Use the local repository as the source of truth: `git status`, `git diff`, task artifacts under `.mavis/team/tasks/<slug>/` when present, and opened files. Do not invent hosting metadata such as labels, CI status, or reviewers.
+
+## Review Order
+
+1. Understand the intent and affected layers.
+2. Check correctness and regressions first.
+3. Verify architecture boundaries: UI does not know DTO/transport, store does not depend on UI, API services do not pull UI/store.
+4. Check public imports through `@/types` and `@/api`, plus barrel updates for new public symbols.
+5. Review UI consistency, styling conventions, and prop shape.
+6. Check tests: unit for mapper/domain logic, behavior tests for HTTP client changes, e2e plan/spec updates for user flows.
+7. **Reuse lens:** flag new helpers that duplicate existing utilities; prefer adjacent shared modules over copy-paste.
+8. **Efficiency lens:** unnecessary work, sequential ops that should be parallel, hot-path bloat, no-op store updates, overly broad reads.
+9. Report validation gaps: lint, stylelint, type-check, or relevant tests not run.
+
+## Output
+
+Lead with findings ordered by severity. Required shape (also in `code-review-mr`):
+
+- Each finding: **`path:line`** — problem — impact — concrete fix direction.
+- Severity: **`blocker` | `important` | `nit`** (list blocker → nit).
+- Vague («looks risky») is not a finding — cite evidence or ask for surrounding context.
+- End with a verdict line: **`Safe to merge | needs changes | reject`**.
+- If there are no findings, say so and call out remaining test or validation risk.
+
+For agent team tasks, write `.mavis/team/tasks/<slug>/review.md` and upsert the owned terminal receipt (`completed` or `changes_requested`); do not mutate `status.json`.
+
+## Receiving review feedback
+
+<!-- shared-core: review/receiving-code-review.md -->
+
+When **implementing** review comments (not when issuing them): verify each item against the codebase before changing code; do not perform agreement (`You're absolutely right!`). Unclear items → ask, do not guess.
+
+## Design guidance
+
+- Load `design-guidance` (rule stem) when assessing structure, smells, or pattern fit.

package/presets/mavis/next/skills/debug-investigation/SKILL.md

@@ -0,0 +1,26 @@
+<!-- shared-core: quality/debug-investigation.md -->
+
+# Debug investigation
+
+Stack-agnostic root-cause loop. Consumed by skill `debug-investigation` and agent `debugger`. Platform paths, MCP, and stack commands belong in adapter **Stack notes**.
+
+## When
+
+Any bug, test failure, build failure, or unexpected behavior — **before** proposing a fix.
+
+## Phases (in order)
+
+1. **Evidence** — reproduce, or collect the closest evidence (error output, failing test, logs, user steps).
+2. **Isolate** — expected vs actual in one sentence; smallest relevant slice of the codebase.
+3. **Root cause** — name the cause before editing. One hypothesis at a time; do not stack speculative fixes.
+4. **Minimal fix + verify** — smallest change that preserves architecture and public contracts. Add a regression test when the behavior is user-facing or likely to recur. Re-run the **failing** check first, then the relevant lint/type/test slice.
+
+## 3-fix architecture gate
+
+If **three** fix attempts fail, or the same symptom returns after two supposed fixes: **STOP**. Do not attempt fix #4. Escalate — the design is likely wrong (human partner or `solution-architect`). Chasing symptoms is the failure mode.
+
+## Do not
+
+- Broad refactor during a bugfix.
+- Fix only where the error surfaces if invalid data originates elsewhere.
+- Claim `fixApplied: true` without re-running the failing check (see `quality/verification-before-completion.md`).

package/presets/mavis/next/skills/feature-delivery/SKILL.md

@@ -0,0 +1,34 @@
+# Feature Delivery
+
+1. Read task brief, acceptance criteria, and decomposition if present.
+2. If `docs/specs/INDEX.md` exists **or** brief/handoff lists approved generated specs, resolve via **`product-specs`** and treat approved specs as behavior SoT before layer work (soft-fail on miss — do not invent behavior). If generation was requested but not yet approved, stop for human gate — do not implement against draft-only gaps.
+3. Read **`feature-delivery-workflow.md`** and **`reference-features.md`** — follow layer order and mirror closest reference feature.
+4. Add or update focused tests for changed behavior.
+5. Validation from `app/`: **invoke** **`post-change-lint.md`** after code edits (requestable, not always-on — but **must** when code changed). Pipeline: scoped lint OK if next step is `build-verifier`.
+
+- If the prompt or brief contains a `figma.com` URL, load requestable skill `mockup-review` and do not start UI work until the mockup-reviewer step is completed or the user continued past FAIL/`BLOCKED_MCP`.
+- When required-signal (user asked UX, AC require flows/states, or spec-gen page/feature) and **no** `figma.com` URL: `ux-design.md` must exist on the analysis manifest entry (verdict not `INCOMPLETE`) before UI work. If missing, **blocked** — hand back to `task-analyst`. Parallel to the mockup gate when Figma exists.
+
+## Scope discipline
+
+Implement only the current AC / decomposition task. Clarify ambiguity before coding; do not add unrelated hardening, polish, or refactors. See `feature-developer` → **Zero improvisation**.
+
+## Handoff
+
+Summarize by layer, validation results, and known gaps. Agent team: upsert the owned terminal receipt in `artifact-manifest.json`; do not mutate `status.json`.
+
+
+## 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.
+
+
+## Design guidance
+
+- Load `design-guidance` (rule stem) when assessing structure, smells, or pattern fit.
+
+## Anti-hallucination
+
+- When editing app TS/TSX, load `react-next-anti-hallucination` (phantom Next/React/RTK imports).

package/presets/mavis/next/skills/mockup-review/SKILL.md

@@ -0,0 +1,24 @@
+# Mockup review
+
+Requestable procedure for agent `mockup-reviewer`. Checklist lives in rule `mockup-review` — do not duplicate it here.
+
+## When to run
+
+- Standalone `/task` intent `mockup-review` (report-only; profile `light`; steps only `mockup-reviewer`).
+- Embedded: `feature` / `refactor` / `spike` / `spec-gen` whose prompt or `brief.md` contains a `figma.com` URL on a UI stack — after `task-analyst` (when present), before architect and developer.
+
+## MCP OAuth preflight
+
+Confirm Figma MCP is configured (see `MCP.md`). Do not ship secrets. If MCP is down or auth fails, write `mockup-review.md` with verdict `BLOCKED_MCP` and receipt `blocked`. Do not invent DS claims. Embedded skip-forward is `/task-continue` (explicit «proceed without MCP»).
+
+## Plugin skills
+
+If plugin skills `figma-design-to-code` / `figma-use` are installed, load them for MCP invocation recipes; do not paste their bodies; do not run design-to-code. Preset owns the checklist, agent, and routing; plugin owns MCP recipes.
+
+## Load the checklist
+
+Load requestable rule `mockup-review` and follow its review dimensions, node caps, and report headings.
+
+## Output
+
+Write `.mavis/team/tasks/<slug>/mockup-review.md` and upsert manifest entry id `mockup-review` (producer `mockup-reviewer`). Receipt mapping is in the agent prompt.

package/presets/mavis/next/skills/playwright-e2e/SKILL.md

@@ -0,0 +1,28 @@
+# Playwright E2E
+
+## Project Layout
+
+- Test root: `app/__tests__/e2e`.
+- Mirror formula: `app/src/<rel>` → `app/__tests__/e2e/<rel>` (drop `app/src`, preserve the relative path).
+- Place artifacts at the deepest mirrored folder of the source under test; page-wide scenarios use the page folder.
+- Test plans: `*.cases.md` files in the mirrored folder.
+- Executable specs: `*.spec.ts` files in the same folder.
+- Seed/setup reference: `app/__tests__/e2e/seed.spec.ts`. Root exceptions: `_shared/`, `smoke.spec.ts`, and `seed.spec.ts`.
+
+Do not introduce top-level `specs/`, `tests/`, or feature-named folders for real coverage unless the user explicitly asks for a sandbox.
+
+## Planner
+
+Create or update the relevant `*.cases.md` file. Each scenario should include title, preconditions, steps, and expected result. Cover happy path, important edge cases, validation, and error states.
+
+## Generator
+
+Generate specs from `*.cases.md` into the same folder. Reuse existing page objects and `_shared/` helpers. Keep test names aligned with scenario names.
+
+## Healer
+
+Fix selectors, waits, setup, or page-object details without weakening business assertions. Use the matching `*.cases.md` file as the behavior contract.
+
+## Validation
+
+Run the affected spec when feasible. If the environment blocks execution, report the command that should be run and the blocker.

package/presets/mavis/next/skills/react-performance/SKILL.md

@@ -0,0 +1,39 @@
+# React performance (next)
+
+On-demand checklist for **implementers** shipping Next.js UI. For a formal diff audit report, use the **`performance-auditor`** agent instead — this skill does not replace it.
+
+## Bundle
+
+- Prefer route-level and component-level code splitting for heavy UI (charts, editors).
+- Avoid importing large libraries into shared layouts; keep client entry points small.
+- Check that barrel files do not re-export heavy modules into every consumer.
+
+## RSC / client boundaries
+
+- Default to Server Components; add `"use client"` only when hooks, browser APIs, or interactivity require it.
+- Keep client leaves small; pass serializable props from server parents.
+- Do not pull server-only modules (secrets, Node APIs) into client components.
+
+## Memoization discipline
+
+- Do **not** add `useMemo` / `useCallback` by default — follow the repo’s React Compiler / team guidance.
+- Memoize when profiling shows unstable props causing expensive subtree re-renders, or when an existing pattern already does.
+- Prefer fixing state placement and list keys over blanket memo wrappers.
+
+## Data-fetch waterfalls
+
+- Avoid sequential awaits that could be parallel (`Promise.all`) when independent.
+- Lift shared fetches to a layout/parent when multiple children need the same data.
+- Prefer store/thunk or server fetch patterns already used in the feature’s reference slice — do not invent a second data layer.
+
+## Core Web Vitals (quick)
+
+| Metric | Implementer checks |
+|--------|--------------------|
+| LCP | Hero/media not blocked by unnecessary client JS; sensible image sizing |
+| INP | Click handlers lean; avoid long sync work on the main thread |
+| CLS | Reserve space for images/skeleton; stable fonts |
+
+## Handoff
+
+If the user asks for a **performance audit report** of a diff/deliverable, route to **`performance-auditor`** (readonly `perf-report.md`), not this skill alone.

package/presets/mavis/next/skills/spec-authoring/SKILL.md

@@ -0,0 +1,21 @@
+# Spec authoring
+
+Requestable — not always-on. Load **`product-specs`** + **`product-specs-authoring`** first.
+
+## Steps
+
+1. **Resolve mode** — Generate (tree exists) vs Bootstrap (create INDEX + templates + first spec). Refuse inventing INDEX without explicit intent/AC/`/spec-start`.
+2. **Ensure templates** — if consumer `docs/specs/_templates/` missing, copy from package SoT `presets/_shared/assets/docs-specs/_templates/` (or seeded init assets).
+3. **Fill frontmatter v1** — all array keys present (may be `[]`); prefer `type: feature` on backend/MCP; default `status: draft`.
+4. **Fill required body headings** — page: Overview, Entry/actors, Primary flows, States, Acceptance criteria; feature: Overview, Shared behavior/constraints, Related pages. Use `N/A` explicitly when needed.
+5. **Update INDEX** in the **same change set** (columns: id, type, domain, title, path, optional routes/aliases/status).
+6. **Self-check** against `product-specs-authoring` checklist; put blockers in `open_questions`.
+7. **Handoff** — list created/updated paths + INDEX diff for human gate (`after:task-analyst`). Approve before feature-developer.
+
+## Sources
+
+Cite AC / user / Figma / labeled code inference. Do not invent product behavior.
+
+## mcp-ts
+
+Analyst-only path (no tech-writer). `routes` = tool names; `ui`/`store` = `[]`.

package/presets/mavis/next/skills/spec-discovery/SKILL.md

@@ -0,0 +1,53 @@
+# Spec discovery
+
+Requestable — **not** always-on. Load **only** in **task-analyst author mode** when the user or AC asks research-before-spec (including `/spec-start` with that intent). Never default consume `/task`. Not a pipeline step.
+
+Load **`product-specs`** (and **`product-specs-authoring`** after this skill if authoring follows). Do **not** invent product behavior. Do **not** create a second spec-file tree.
+
+## Load policy
+
+1. Pick a **subset** of methods that fit the task — not all 11 every time.
+2. **`claim-ledger` is mandatory** whenever this skill runs.
+3. Write chronological notes to `.mavis/team/tasks/<slug>/.memlog.md` (append-only). Do **not** mutate `status.json` / `metrics.json`.
+4. Discovery output lives in memlog + `brief.md` sections. Then hand off to skill **`spec-authoring`** (or stop if `specMode: none`).
+
+Backend / MCP stacks: skip UI-only examples; prefer `analog-scan`, `code-archaeology`, `problem-define`, `assumption-audit`, `claim-ledger`.
+
+## Methods
+
+| Method | Produces | Skip when |
+|--------|----------|-----------|
+| **intake-capture** | Raw idea, origin, first unknowns → memlog + brief Research inputs | User prompt + AC already complete |
+| **index-locator** | Related INDEX rows + linked spec paths + gaps | `specMode: none` or INDEX absent without consume intent |
+| **code-archaeology** | Observed behavior labeled `inferred`/`draft` with path citations | Greenfield with no relevant code |
+| **research-pack-scan** | Claims from `team/tasks/<slug>/research/**` or user-supplied paths | No research paths cited |
+| **evidence-against** | Reasons the idea may fail | Internal spike with no product/external risk |
+| **problem-define** | Who hurts, goals, non-goals, cost of inaction → brief | Brief Goal + Out of scope already exhaustive |
+| **concept-shape** | 2–3 approach options + appetite (small/medium/large); no API design | `architecture.md` exists or one obvious path |
+| **working-backwards-check** | Headline + 5 customer objections + verdict `forged` / `needs-heat` / `cracked` | Consume mode with approved specs |
+| **assumption-audit** | Assumption register → brief Assumptions + `open_questions` | Never skip in author mode (keep lightweight) |
+| **analog-scan** | 2–3 repo analogs via `reference-features` / code search | No plausible analog |
+| **claim-ledger** | Markdown table `source → claim` for every product-behavior claim | **Never skip** when this skill is loaded |
+
+## Claim ledger (required)
+
+Block spec authoring until every product-behavior claim has a source:
+
+```markdown
+| Source | Claim |
+|--------|-------|
+| brief AC-3 | Checkout shows empty cart copy |
+| code-archaeology inferred | Empty state uses an illustration |
+```
+
+Sources: AC / user / UI copy / labeled code inference / research path. Unsourced claims → `open_questions`.
+
+## Memlog
+
+Path: `.mavis/team/tasks/<slug>/.memlog.md`. Chronological `## YYYY-MM-DD` headings. Optional tags: `decision`, `assumption`, `override`, `event`, `claim`.
+
+## Forbidden
+
+- 108-technique catalogs, HTML composers, party-mode, five `assess.*` slash commands, Python memlog runtime.
+- Dummy INDEX rows when `specMode: none`.
+- Dumping the whole `docs/specs/**` tree — locator is INDEX → linked page + optional `feature.md`.

package/presets/mavis/next/skills/technical-retro/SKILL.md

@@ -0,0 +1,49 @@
+# Technical Retro
+
+## Inputs
+
+Ask for missing period, focus, task slug, participants, or format constraints. If a slug is provided, read `.mavis/team/tasks/<slug>/` artifacts and local git status/diff. Do not invent facts not present in artifacts or the user's message.
+
+## Facilitation Principles
+
+- Focus on process and system behavior, not blame.
+- Separate facts, interpretations, and proposed actions.
+- Balance what worked with what slowed delivery.
+- Keep action items concrete, owned, and measurable.
+
+## Default Structure
+
+1. Goal and scope.
+2. Timeline or factual observations.
+3. What worked.
+4. What hurt quality, speed, or confidence.
+5. Root causes for the top one or two issues.
+6. Experiments and action items.
+
+## Agent Team Block
+
+For `/task` pipelines, include router, analyst, developer, **build-verifier**, reviewer, QA/e2e, hooks, and rules.
+
+## Wrong-turn taxonomy (optional)
+
+When artifacts or the user mention agent mistakes, tag incidents:
+
+| Kind | Signal |
+|------|--------|
+| `correction` | User pushback («no», «don't», «actually») |
+| `retry` | Same tool/approach 2+ times before success |
+| `waste` | Many searches before the right file |
+| `reversal` | Edit then revert / delete |
+| `dead-end` | Env/tool failure (missing binary, wrong path) |
+
+Severity: **high** (explicit correction / large waste) · **med** · **low**. Confidence 0–100 — be honest; low-confidence items are optional.
+
+## Preset feedback loop
+
+If the same issue appeared in **2+** tasks (or **3+** for packaging a new rule/skill — Rule of Three), add section **Preset updates**:
+
+- Proposed change to `.mavis/rules/`, `.mavis/agents/`, `.mavis/skills/`, or **`.mavis/team/conventions.md`** (prefer conventions for project gotchas)
+- One concrete snippet or file path
+- Whether it belongs in always-on vs globs vs skill vs conventions (see `PRESET-CONTRIBUTION` → Building blocks)
+
+Do not apply preset or conventions changes without user approval.

package/presets/mavis/next/skills/unit-testing/SKILL.md

@@ -0,0 +1,29 @@
+# Unit Testing
+
+## Project Layout
+
+- Test root: `app/__tests__/unit`.
+- Mirror formula: `app/src/<rel>` → `app/__tests__/unit/<rel>` (drop `app/src`, preserve the relative path).
+- Place specs at the deepest mirrored folder of the module under test; do not colocate tests under `app/src/**`.
+- Legacy `*.test.ts` / `*.test.tsx` — do not rename without a separate task.
+- Runner: Jest or Vitest as configured in the target repo; components use `@testing-library/react`.
+
+Follow **`tests-unit.md`** (Cursor) or **`rules/testing/tests-unit.md`** (Claude).
+
+## Planner
+
+Write or update `.mavis/team/tasks/<slug>/unit-test-plan.md` (or `.claude/team/...`).
+
+Each scenario needs a Russian `it()` title (capital letter after each sentence). Cover mappers, thunks, client behavior branches, and component interactions per AC. Record the mirrored target path under `app/__tests__/unit/`.
+
+## Generator
+
+Implement specs from `unit-test-plan.md` under the mirrored `app/__tests__/unit/` path. Import the source under test through the `@/` alias. Reuse project test-store, HTTP mocks, and existing test utilities. Do not change production code.
+
+## Healer
+
+Fix mocks, async timing, selectors, and setup — preserve business assertions. Use `unit-test-plan.md` and `brief.md` as the behavior contract.
+
+## Validation
+
+Run affected unit tests from `app/` when feasible. Report command, pass/fail, and remaining gaps vs plan.

package/presets/mavis/next/skills/ux-design/SKILL.md

@@ -0,0 +1,31 @@
+# UX design
+
+Requestable procedure for **task-analyst** (not a pipeline agent). Checklist lives in rule `ux-design` — do not duplicate it here.
+
+**Not** a Figma workflow: do not call Figma MCP; do not write `DESIGN.md`. Visual/DS/viewport audit stays with `mockup-reviewer` / skill `mockup-review`.
+
+## When to run
+
+- Standalone `/task` intent `ux-design` (report-only; profile `light`; steps only `task-analyst`; modes `synthesize` | `review`; `humanGates: []`).
+- Embedded: `feature` / `refactor` / `spike` / `spec-gen` on a UI stack when required-signal or advisory UI AC fires — inside the analyst step (no new pipeline agent). Required-signal: user asked UX, or AC require flows/states, or spec-gen page/feature.
+
+## Modes
+
+- `synthesize` — elicit IA, named primary flows, empty/error/loading matrix from the brief/AC.
+- `review` — critique a brief/wireframe/description **without** Figma. A `figma.com` URL with review-not-implement language is intent `mockup-review`, not this skill.
+
+## Order
+
+UX synthesis **before** skill `spec-authoring` / author mode when both apply. Analyst remains the only INDEX writer. Map headings via the rule's Spec feed table; do not paste the whole UX file into `docs/specs`.
+
+## Forbidden Figma MCP
+
+Do not call `get_metadata`, `get_design_context`, `get_screenshot`, `search_design_system`, `get_variable_defs`, `use_figma`, `create_new_file`, `generate_*`. Only `mockup-reviewer` reads Figma.
+
+## Load the checklist
+
+Load requestable rule `ux-design` and follow its required-signal vs advisory rules, forbidden MCP list, verdicts, surface-closure, and report headings.
+
+## Output
+
+Write `.mavis/team/tasks/<slug>/ux-design.md` and append that path to the **`analysis`** manifest entry (`authoritativePaths` / `evidencePaths`). Do **not** create a new producer row. Receipt stays on `analysis`: required `INCOMPLETE` (missing primary flows or empty/error/loading matrix) → DoR not met; advisory `INCOMPLETE` → notes only. Never copy the mockup FAIL-only hook.

package/presets/mavis/next/skills/write-adr/SKILL.md

@@ -0,0 +1,36 @@
+# Write ADR
+
+On-demand skill for durable architecture decisions. Stack-agnostic sections — fill with project-specific paths and links.
+
+## When to use
+
+- Choosing between non-trivial alternatives (storage, layering, library, API shape).
+- Recording a decision that future agents/humans should not silently reverse.
+- User asks for an ADR, decision log, or “why did we pick X”.
+
+## Template
+
+Create (or append) a short markdown ADR in the repo’s agreed location (e.g. `docs/adr/`, `.mavis/team/`, wiki). Suggested filename: `NNNN-short-title.md`.
+
+```markdown
+# ADR NNNN: <Title>
+
+## Context
+What problem / constraint / force requires a decision?
+
+## Decision
+What we chose (one clear option). Include rejected alternatives briefly if useful.
+
+## Consequences
+Positive and negative outcomes; follow-up work; what becomes harder.
+
+## Links
+Related PRs, issues, rules, prior ADRs, diagrams.
+```
+
+## Guidelines
+
+- Prefer one decision per ADR.
+- Keep Context factual; Decision imperative and short.
+- Link living gotchas to `team/conventions.md` when the decision becomes a recurring local pattern.
+- Do not put ADR bodies only in root `AGENTS.md` (overwrite on init).