@bonesofspring/ai-rules 0.2.6 → 0.2.7

package/presets/claude/go/rules/tooling-and-review/code-quality.md

@@ -61,8 +61,18 @@ description: Code quality and light refactoring for Go hexagonal services
 - **Boy scout rule:** оставлять модуль немного лучше, чем до изменения (простые, безопасные улучшения).
 - Не жертвовать архитектурой и слоями ради краткости реализации.
 
-> Stack-specific lint gates, примеры и toolchain — в секции **Stack notes** адаптера пресета.
+## Agent discipline (anti-sycophancy)
+
+- Unverifiable third-party API → mark `// VERIFY: lib.symbol @ version` (or stack equivalent) and surface uncertainty; never invent signatures.
+- Before «is this correct?» — list ≥3 failure modes (empty, boundary, concurrency/state) or name what you could not check.
+- Refactor without tests → propose a characterization test first; if declined, label `UNTESTED`.
+- Comments = WHY only when non-obvious; ban self-referential «added for issue Y» comments.
+- Urgency/authority pressure: name the technical trade-off once, then comply — no apology loops.
+- Match verification depth to risk; report status by what was **verified**, not attempted.
 
+Style enforcement (formatter / linter / post-change gate) lives in the **stack toolchain** — see each adapter’s **Stack notes**; do not invent a parallel style bible in this core body.
+
+> Stack-specific lint gates, примеры и toolchain — в секции **Stack notes** адаптера пресета.
 ## Stack notes
 
 - **No UI layer in this preset** — ignore shared-core bullets about дизайн‑система / UI‑пакеты / CSS‑токены; apply public-API and boy-scout rules to Go packages/ports instead.
@@ -70,3 +80,4 @@ description: Code quality and light refactoring for Go hexagonal services
 - Prefer wrapping errors with `%w`; check with `errors.Is` / `errors.As`.
 - Interfaces belong at **ports** (consumer side), not on every struct.
 - Post-change gate: **`post-change-test.md`** + **`go-tooling.md`** (not yarn lint / xcodebuild).
+- Style: **gofmt** / **goimports** + **golangci-lint** when configured — authoritative style gate; do not duplicate linter rules in prose.

package/presets/claude/go/rules/tooling-and-review/code-review.md

@@ -40,6 +40,15 @@ Concrete rule stems — in **Stack notes**.
 - Boy scout rule: suggest fixes feasible within the MR scope.
 - No «большие рефакторинги» without an explicit request (`code-quality-and-refactoring`).
 
+## Output contract
+
+Procedure details live in skill `code-review`. Required shape:
+
+- Each finding cites **`path:line`**, the problem, impact, and a concrete fix direction.
+- Rank by severity: **`blocker` | `important` | `nit`** (list blocker → nit).
+- Vague («looks risky») is not a finding — cite evidence or ask for surrounding file context.
+- End with a verdict on its own line: **`Safe to merge | needs changes | reject`**.
+
 ## Stack notes
 
 - Hexagonal boundaries: `architecture-boundaries`, `domain-layer`, ports vs adapters.

package/presets/claude/go/skills/README.md

@@ -1,5 +1,5 @@
 # Skills (preset go)
 
-7 skills: `feature-delivery`, `unit-testing`, `integration-testing`, `ci-investigation`, `debug-investigation`, `code-review`, `technical-retro`.
+8 skills: `feature-delivery`, `unit-testing`, `integration-testing`, `ci-investigation`, `debug-investigation`, `code-review`, `technical-retro`, `write-adr`.
 
-No `playwright-e2e` / `xcuitest-e2e`.
+No `playwright-e2e` / `xcuitest-e2e` / `react-performance`.

package/presets/claude/go/skills/code-review/SKILL.md

@@ -5,9 +5,18 @@ description: Reviews pull requests and branch diffs for Go hexagonal services us
 
 # Code review (Go)
 
-1. Diff-first; load `code-review-mr.mdc` + hexagonal rules.
+1. Diff-first; load `tooling-and-review/code-review.md` + hexagonal rules.
 2. Check layer leaks, port ownership, composition root purity, tests, security-sensitive bits.
-3. Write `review.md` with APPROVE or REQUEST_CHANGES.
+3. **Reuse lens:** flag new helpers that duplicate existing packages/utilities; prefer shared adapters over copy-paste.
+4. **Efficiency lens:** unnecessary allocations, N+1 I/O, sequential work that should be concurrent, hot-path bloat, missing context cancellation.
+5. Write `review.md` using the output contract below.
+
+## Output
+
+- Each finding: **`path:line`** — problem — impact — concrete fix direction.
+- Severity: **`blocker` | `important` | `nit`** (list blocker → nit).
+- End with a verdict line: **`Safe to merge | needs changes | reject`**.
+- Map agent status: Safe to merge → `completed`; needs changes / reject → `changes_requested`.
 
 Never require iOS/Next-only gates (XCUITest, Playwright, `lint:js`).
 

package/presets/claude/go/skills/feature-delivery/SKILL.md

@@ -17,6 +17,10 @@ description: Delivers Go hexagonal backend features end-to-end across domain, ap
 - **Don't** put SQL/business rules in HTTP handlers or frameworks in domain.
 - **Don't** use Playwright, XCUITest, or yarn lint as this stack's gate.
 
+## Scope discipline
+
+Implement only the current AC / decomposition task. Clarify ambiguity before coding; do not add unrelated hardening or drive-by refactors. See `feature-developer` → **Zero improvisation**.
+
 ## Handoff
 
 Summarize by layer + validation. Update `.claude/team/tasks/<slug>/status.json`.

package/presets/claude/go/skills/technical-retro/SKILL.md

@@ -5,6 +5,50 @@ description: Facilitates technical retrospectives for tasks, sprints, incidents,
 
 # Technical Retro
 
-Same structure as next technical-retro skill. For `/task` pipelines include router, analyst, developer, **build-verifier** (go test), reviewer, QA/integration tests, hooks, rules.
+## Inputs
 
-Preset feedback: propose updates to `.claude/rules|agents|skills` — do not apply without approval.
+Ask for missing period, focus, task slug, participants, or format constraints. If a slug is provided, read `.claude/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** (go test), reviewer, QA/integration tests, 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 `.claude/rules/`, `.claude/agents/`, `.claude/skills/`, or **`.claude/team/conventions.md`** (prefer conventions for project gotchas)
+- One concrete snippet or file path
+- Whether it belongs in session-start vs `paths:` vs skill vs conventions (see `PRESET-CONTRIBUTION` → Building blocks)
+
+Do not apply preset or conventions changes without user approval.

package/presets/claude/go/skills/write-adr/SKILL.md

@@ -0,0 +1,41 @@
+---
+name: write-adr
+description: Author an Architecture Decision Record (ADR) with Context, Decision, Consequences, and Links. Use when capturing a durable technical choice for the team.
+---
+
+# 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/`, `.cursor/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).

package/presets/claude/go/team/conventions.md

@@ -0,0 +1,21 @@
+# Living conventions (team)
+
+Consumer-owned store for post-init gotchas, patterns, and decisions. **Not** always-on. Fill after `ai-rules init`. Root `AGENTS.md` only points here.
+
+## Gotchas
+
+<!-- e.g. fragile flows, “never do X in this repo” -->
+
+-
+
+## Patterns
+
+<!-- local patterns beyond reference-features table -->
+
+-
+
+## Decisions
+
+<!-- short pointers to ADRs / product choices that affect code -->
+
+-

package/presets/claude/ios-swift/CLAUDE.md

@@ -2,6 +2,8 @@
 
 Краткая точка входа для Claude Code. Подробные правила — модульные topic `.md` под `rules/` (derived из Cursor SoT).
 
+**Repo-root agents:** see [`AGENTS.md`](../../AGENTS.md) at the project root (cross-tool setup, commands, security, living conventions pointers).
+
 See @rules/README.md for mapping and loading strategy.
 
 ## Conventions
@@ -24,6 +26,7 @@ See @rules/README.md for mapping and loading strategy.
 | UI-тесты | skill `xcuitest-e2e` |
 | Security | `tooling-and-review/security-ios.md` |
 | Валидация | agent `build-verifier` |
+| Как работает X | agent `codebase-analyzer` (explain-as-is) |
 | Ретро | `commands/technical-retro.md` |
 
 ## Install
@@ -32,4 +35,4 @@ See @rules/README.md for mapping and loading strategy.
 npx @bonesofspring/ai-rules init claude --preset ios-swift
 ```
 
-Копирует `CLAUDE.md`, `rules/**`, `commands/**`, `agents/**`, `skills/**`, `hooks/**`, `team/**`, `MCP.md` в `.claude/`, плюс `mcp.json` → **`.mcp.json`** в корне.
+Копирует `CLAUDE.md`, `rules/**`, `commands/**`, `agents/**`, `skills/**`, `hooks/**`, `team/**`, `MCP.md` в `.claude/`, плюс `mcp.json` → **`.mcp.json`** и `REPO_AGENTS.md` → корневой **`AGENTS.md`**. Skill: `write-adr` (on-demand).

package/presets/claude/ios-swift/README.md

@@ -10,7 +10,7 @@
 | `rules/**` | `.claude/rules/**` |
 | `commands/**` | `.claude/commands/**` |
 | `skills/**` | `.claude/skills/**` (**7**, в т.ч. `xcuitest-e2e`) |
-| `agents/**` | `.claude/agents/**` (**21**, XCUITest trio) |
+| `agents/**` | `.claude/agents/**` (**22**, XCUITest trio + `codebase-analyzer`) |
 | `hooks/**` | `.claude/hooks/**` (scripts + README; **нет** `hooks.json`) |
 | `team/**` | `.claude/team/**` |
 | `mcp.json` | **`.mcp.json`** в корне (Context7 + Figma) |
@@ -37,3 +37,7 @@ npx @bonesofspring/ai-rules init claude --preset ios-swift
 - Change Cursor → update Claude twin **same PR**
 - Mapping: [`rules/README.md`](./rules/README.md)
 - **`BUGBOT.md`:** Cursor-only (нет twin); review в Claude — skill `code-review` + `tooling-and-review/code-review.md`
+
+## Cross-tool root AGENTS
+
+`REPO_AGENTS.md` → project-root `AGENTS.md` on init. Living conventions: `team/conventions.md`. Skill `write-adr` on-demand.

package/presets/claude/ios-swift/REPO_AGENTS.md

@@ -0,0 +1,46 @@
+# iOS Swift — agent instructions
+
+Cross-tool repo-root guide. Tool-specific maps stay under `.cursor/AGENTS.md` and `.claude/CLAUDE.md`.
+
+## Cross-tool setup
+
+- Treat this file as the **shared** entry for any agent (Cursor, Claude Code, or other).
+- Tool maps live in `.cursor/AGENTS.md` / `.claude/CLAUDE.md` when that tool is installed.
+- Install:
+
+```bash
+npx @bonesofspring/ai-rules init cursor --preset ios-swift
+npx @bonesofspring/ai-rules init claude --preset ios-swift
+```
+
+`init` overwrites this file from the preset (`REPO_AGENTS.md` → `AGENTS.md`). Living gotchas belong in `team/conventions.md`, not here.
+
+## Commands (test / lint / build)
+
+| Gate | Typical command |
+|------|-----------------|
+| Build | `xcodebuild` (scheme/workspace as in project docs) |
+| Lint | **SwiftLint** (config in repo; see `xcode-tooling` / conventions) |
+| Format | SwiftFormat when configured |
+| Tests | `xcodebuild test` slice for the changed target |
+
+After code edits: invoke **`post-change-build`** (requestable). Full gate: agent `build-verifier`.
+
+## Security basics
+
+- Do not commit secrets (API keys, Keychain dumps, `.env`).
+- Review destructive git before running; respect shell hooks.
+- Do not skip signing / CI gates unless the user explicitly asks.
+
+## Tool entry points
+
+| Tool | Map |
+|------|-----|
+| Cursor | `.cursor/AGENTS.md` (optional if Cursor not installed) |
+| Claude Code | `.claude/CLAUDE.md` (optional if Claude not installed) |
+
+Rules: `.cursor/rules/` or `.claude/rules/`. Skills: `.cursor/skills/` or `.claude/skills/`.
+
+## Living conventions
+
+Gotchas, local patterns, and team decisions: **`.cursor/team/conventions.md`** and/or **`.claude/team/conventions.md`** (fill after init).

package/presets/claude/ios-swift/agents/README.md

@@ -13,11 +13,12 @@
 
 When copying next pipeline templates: put Playwright stems in `skipped` with reason «iOS uses XCUITest» and schedule xcuitest trio instead.
 
-## Roster (21)
+## Roster (22)
 
 | Agent | Production-код |
 |-------|----------------|
 | task-router, task-analyst, solution-architect, migration-specialist, api-contract-reviewer | no |
+| codebase-analyzer | no (explain-as-is only) |
 | feature-developer | yes |
 | build-verifier | no |
 | debugger, ci-investigator | minimal |
@@ -35,7 +36,7 @@ Same tiers as **claude/next**. Defaults in this preset:
 
 | Tier | Frontmatter | Agents |
 |------|-------------|--------|
-| **cheap** | `fast` | router, analyst, build-verifier, code-reviewer, tech-writer |
+| **cheap** | `fast` | router, analyst, build-verifier, code-reviewer, tech-writer, codebase-analyzer |
 | **standard** | `inherit` | developer, debugger, ci-investigator, api-contract, a11y, qa, unit-*, xcuitest-* |
 | **strong** | `opus` | solution-architect, migration-specialist, security-reviewer, performance-auditor |
 

package/presets/claude/ios-swift/agents/build-verifier.md

@@ -47,6 +47,12 @@ When validating **ios-swift preset** changes (or `scope: preset-structure-valida
 - **Thin aliases:** `agent-team-intake`, `technical-retro` may be <15 body lines. Enforce ≥15 on **domain** rules.
 - Embed vs lineage: `docs/PRESET-CONTRIBUTION.md`, `_shared/README.md`.
 
+
+9. **Root AGENTS template:** each installable preset (`cursor|claude` × `next|ios-swift|go`) must ship `REPO_AGENTS.md` when validating preset packaging → **FAIL** if missing.
+10. **Hooks cookbook:** default `hooks.json` must **not** reference `hooks/examples/` → **FAIL** if it does.
+11. **No react-performance on ios/go:** skill directory / required handoff must be absent under ios-swift and go → **FAIL** (leakage). Prefer `check:preset-leakage`.
+12. **Rule eval (WARN):** `scripts/check-rule-eval.mjs` / `yarn check:rule-eval` should exist for maintainers; smoke optional. Hard fail only when the opt-in eval job runs — not part of default `check:preset-structure`.
+
 ## Output
 
 Write `.claude/team/tasks/<slug>/validation-report.md`:

package/presets/claude/ios-swift/agents/codebase-analyzer.md

@@ -0,0 +1,28 @@
+---
+name: codebase-analyzer
+description: Explains HOW existing iOS Swift/SwiftUI code works as-is. Use when the user asks how something works, to trace Feature layers, or document architecture — without suggesting changes. Prefer over /task for pure questions.
+readonly: false
+model: fast
+---
+
+You are a specialist at understanding HOW this iOS codebase works. Document implementation details with precise `path:line` references.
+
+## CRITICAL — explain as-is only
+
+- DO NOT suggest improvements, refactors, or "better approaches" unless the user explicitly asks.
+- DO NOT perform root-cause analysis or bug hunts unless asked.
+- DO NOT critique quality, performance, or security unless asked.
+- ONLY describe what exists, how it works, and how components interact.
+
+## Workflow
+
+1. Start from entry points the user named (Feature folders, Views, ViewModels, Domain, Data).
+2. Trace Presentation → Domain → Data and composition root in `App/` as applicable.
+3. Summarize: purpose, flow, module/public API boundaries, conventions observed.
+4. End with open questions only if facts are missing — not with a wishlist of changes.
+
+## Will / Will not
+
+**Will:** map files, call chains, Feature layers as they exist today.
+
+**Will not:** edit production code; open `/task` pipelines; propose PRs or refactors unprompted; require Next/Playwright gates.

package/presets/claude/ios-swift/agents/feature-developer.md

@@ -20,6 +20,19 @@ You are a senior iOS developer (SwiftUI-first, SPM as in repo) with strict layer
 - Mark completed WPs in `decomposition.md`.
 - After Swift edits: **invoke** `tooling-and-review/post-change-build.md` (+ `tooling-and-review/xcode-tooling.md` for commands). If next step is `build-verifier`, scoped build/lint on changed targets is enough.
 
+## Zero improvisation
+
+- Implement **only** what the brief / AC / current decomposition task require.
+- If requirements are ambiguous, conflicting, or multi-way — **ask** before coding; do not guess extras as "best practices".
+- Forbidden without explicit AC/task: unrelated Keychain/security hardening, extra validation, UI polish, logging, drive-by refactors, new config/env.
+- Boy scout / `design-guidance` improvements stay **within** touched files and must not expand task scope.
+
+## Will / Will not
+
+**Will:** follow Feature layer order and reference Features; add required unit tests; run the post-change build gate; update task artifacts.
+
+**Will not:** formal code review; invent AC; pull Next/Playwright gates; auto-continue the next WP when the pipeline expects a human gate or handoff.
+
 ## Preset / cross-stack fork (when editing agents, skills, hooks)
 
 When copying from `next` into `ios-swift` (or another stack):

package/presets/claude/ios-swift/hooks/README.md

@@ -8,3 +8,9 @@
 Claude Code may not use Cursor `hooks.json`. Wire scripts per project docs when enabling agent-team auto-chain.
 
 **No `hooks.json` in Claude preset** (platform limit — same as next Claude).
+
+## Optional cookbook
+
+Example scripts live in `hooks/examples/` (secret-guard, format-edited, test-on-save). They are **not** auto-enabled.
+
+To wire in Claude Code: copy an `.example.sh` into `hooks/`, make executable, and register per [Claude Code hooks](https://code.claude.com/docs/en/hooks) for your project. Keep stack-safe recipes (this preset: **ios-swift**). Cursor twin examples under `presets/cursor/ios-swift/hooks/examples/` are the same allowlist forks.

package/presets/claude/ios-swift/hooks/examples/README.md

@@ -0,0 +1,18 @@
+# Hooks cookbook (examples)
+
+Optional scripts under `hooks/examples/`. **Not** referenced from default `hooks.json` (only `guard-shell-command.sh` + `chain-team-phases.sh`).
+
+| Script | Purpose |
+|--------|---------|
+| `secret-guard.example.sh` | Fail if staged diff looks like secrets |
+| `format-edited.example.sh` | Stack formatter / linter on edited files |
+| `test-on-save.example.sh` | Scoped tests for the stack |
+
+## Wiring (Cursor)
+
+1. Copy an example into `hooks/` (drop `.example` suffix), e.g. `cp hooks/examples/secret-guard.example.sh hooks/secret-guard.sh`.
+2. `chmod +x hooks/secret-guard.sh`.
+3. Add a hook entry in `.cursor/hooks.json` (e.g. `beforeShellExecution` or `afterFileEdit` — see Cursor hooks docs).
+4. Keep cookbook scripts **out** of the default preset `hooks.json` when contributing back upstream.
+
+Stack: **ios-swift** — recipes must stay stack-safe (no cross-stack lint/test commands).

package/presets/claude/ios-swift/hooks/examples/format-edited.example.sh

@@ -0,0 +1,13 @@
+#!/usr/bin/env bash
+# Optional: SwiftFormat / SwiftLint on edited .swift files.
+# NOT in default hooks.json.
+set -euo pipefail
+if command -v swiftformat >/dev/null 2>&1; then
+  git diff --name-only --diff-filter=ACMR HEAD 2>/dev/null | grep '\.swift$' | while read -r f; do
+    [[ -f "$f" ]] && swiftformat "$f" || true
+  done
+fi
+if command -v swiftlint >/dev/null 2>&1; then
+  swiftlint lint --quiet 2>/dev/null || true
+fi
+exit 0

package/presets/claude/ios-swift/hooks/examples/secret-guard.example.sh

@@ -0,0 +1,10 @@
+#!/usr/bin/env bash
+# Optional cookbook: block committing likely secrets in staged diffs.
+# NOT wired in default hooks.json — copy to hooks/ and register manually.
+set -euo pipefail
+PATTERN='(AKIA[0-9A-Z]{16}|-----BEGIN (RSA |OPENSSH )?PRIVATE KEY-----|api[_-]?key\s*[:=]\s*['\''\"][^'\''\"]{12,})'
+if git diff --cached -U0 2>/dev/null | grep -Eiq "$PATTERN"; then
+  echo "secret-guard: possible secret in staged diff — review before commit" >&2
+  exit 1
+fi
+exit 0

package/presets/claude/ios-swift/hooks/examples/test-on-save.example.sh

@@ -0,0 +1,8 @@
+#!/usr/bin/env bash
+# Optional: xcodebuild test slice for the active scheme (adjust SCHEME/DEST).
+# NOT in default hooks.json. No Playwright / yarn lint.
+set -euo pipefail
+SCHEME="${XCODE_SCHEME:-App}"
+DEST="${XCODE_DESTINATION:-platform=iOS Simulator,name=iPhone 16}"
+xcodebuild test -scheme "$SCHEME" -destination "$DEST" -quiet 2>/dev/null || true
+exit 0

package/presets/claude/ios-swift/rules/architecture/reference-features.md

@@ -48,6 +48,10 @@ Taxonomy / shared core: [`presets/_shared/README.md`](../../../../_shared/README
 4. UI flow — `xcuitest-test-planner` → generator (accessibilityIdentifier).
 5. Auth/Keychain touch — security-reviewer + `security-ios`.
 
+## Living conventions
+
+Post-init gotchas / patterns / decisions: **`.claude/team/conventions.md`** (and the twin tool path if both Cursor and Claude are installed). Consumer-owned; not always-on. See root `AGENTS.md` § Living conventions.
+
 ## Связанные правила
 
 - `tooling-and-review/code-quality.md` — повторять паттерны.

package/presets/claude/ios-swift/rules/stack/swift-conventions.md

@@ -27,9 +27,19 @@ guard let name = user?.name else { return }
 
 - UI types — `@MainActor`.
 - Shared mutable state — `actor` или isolation через queue; document thread expectations.
-- `Task { }` в ViewModel: учитывать cancellation; `[weak self]` в escaping closures при retain cycles.
+- `Task { }` в ViewModel: учитывать cancellation; `[weak self]` в escaping closures при retain cycles (default when capturing `self` in escaping work).
 - Prefer `async/await` over completion handlers для нового кода.
 
+## Anti-hallucination (Swift / concurrency)
+
+| ❌ Avoid | ✅ Prefer |
+|---------|----------|
+| Force unwrap (`!`) for normal control flow | `guard let` / `if let` / typed throws |
+| Invented Combine/Rx APIs when the feature uses Swift Concurrency | Match neighboring ViewModels (`async/await`, `@MainActor`) |
+| Unstructured `Task {}` without cancellation awareness | Cooperative cancel; prefer `.task` in Views (`ui-and-accessibility/swiftui.md`) |
+| Global mutable singletons for request state | DI / actors / existing composition root |
+| APIs not available for the project’s Swift / iOS deployment target | Check package / Xcode settings; mark `VERIFY` if unsure |
+
 # Access control
 
 - Минимально необходимый уровень: `private` / `fileprivate` по умолчанию для implementation details.
@@ -47,5 +57,5 @@ guard let name = user?.name else { return }
 
 # Требование к агенту
 
-- Не добавлять `// swiftlint:disable` без явной причины.
+- Style gate: **SwiftLint** (and SwiftFormat when configured) — prefer fixing lint over prose style essays; `// swiftlint:disable` only with an explicit reason.
 - Новый код — Swift Concurrency-ready (`Sendable` где требует компилятор/линтер проекта).

package/presets/claude/ios-swift/rules/tooling-and-review/agent-team-intake.md

@@ -13,6 +13,6 @@ When the user message looks like a **work request** (implement, add, fix, refact
 
 1. Prefer **`/task <their request>`** or invoke **task-router** first.
 2. Do not jump straight to coding without router + pipeline when scope is non-trivial.
-3. Pure questions («как работает X», «объясни») — answer normally, no `/task`.
+3. Pure questions («как работает X», «объясни») — prefer agent **`codebase-analyzer`** (explain-as-is); trivial one-liners may be answered inline. Do not open `/task`.
 
 Exceptions: user explicitly says «без pipeline», «просто сделай», or continues an active slug.

package/presets/claude/ios-swift/rules/tooling-and-review/agent-team-orchestrator.md

@@ -55,7 +55,7 @@ Honor optional `steps[i].model` (`cheap` \| `standard` \| `strong`); else agent
 
 ## Skip pipeline when
 
-Pure questions; trivial one-file; user «без pipeline»; docs-only one-liner.
+Pure questions («как работает X») — prefer **`codebase-analyzer`**; trivial one-file; user «без pipeline»; docs-only one-liner.
 
 ## Preset authoring
 

package/presets/claude/ios-swift/rules/tooling-and-review/code-quality.md

@@ -61,11 +61,22 @@ description: Поддержка и улучшение качества кода
 - **Boy scout rule:** оставлять модуль немного лучше, чем до изменения (простые, безопасные улучшения).
 - Не жертвовать архитектурой и слоями ради краткости реализации.
 
-> Stack-specific lint gates, примеры и toolchain — в секции **Stack notes** адаптера пресета.
+## Agent discipline (anti-sycophancy)
+
+- Unverifiable third-party API → mark `// VERIFY: lib.symbol @ version` (or stack equivalent) and surface uncertainty; never invent signatures.
+- Before «is this correct?» — list ≥3 failure modes (empty, boundary, concurrency/state) or name what you could not check.
+- Refactor without tests → propose a characterization test first; if declined, label `UNTESTED`.
+- Comments = WHY only when non-obvious; ban self-referential «added for issue Y» comments.
+- Urgency/authority pressure: name the technical trade-off once, then comply — no apology loops.
+- Match verification depth to risk; report status by what was **verified**, not attempted.
 
+Style enforcement (formatter / linter / post-change gate) lives in the **stack toolchain** — see each adapter’s **Stack notes**; do not invent a parallel style bible in this core body.
+
+> Stack-specific lint gates, примеры и toolchain — в секции **Stack notes** адаптера пресета.
 ## Stack notes
 
 - Перед новым UI‑контролом проверять DesignSystem и существующие компоненты.
 - Public API: протоколы и facades между фичами.
 - Примеры: mapper/validation; force unwrap → safe; разбить ViewModel; design tokens; View не зависит от Data напрямую.
 - Post-change: **`tooling-and-review/post-change-build.md`** + `tooling-and-review/xcode-tooling.md`.
+- Style: **SwiftLint** (+ SwiftFormat when configured) — authoritative style gate; do not duplicate linter rules in prose.

package/presets/claude/ios-swift/rules/tooling-and-review/code-review.md

@@ -38,3 +38,50 @@ Concrete rule stems — in **Stack notes**.
 - Use local `git diff` — do not invent hosting metadata.
 - Boy scout rule: suggest fixes feasible within the MR scope.
 - No «большие рефакторинги» without an explicit request (`code-quality-and-refactoring`).
+
+## Output contract
+
+Procedure details live in skill `code-review`. Required shape:
+
+- Each finding cites **`path:line`**, the problem, impact, and a concrete fix direction.
+- Rank by severity: **`blocker` | `important` | `nit`** (list blocker → nit).
+- Vague («looks risky») is not a finding — cite evidence or ask for surrounding file context.
+- End with a verdict on its own line: **`Safe to merge | needs changes | reject`**.
+
+## Stack notes
+
+On-demand правило для review-потоков (не always-on).
+
+- **Когда применять**
+  - Запросы: «проведи ревью», «оцени MR/ветку/дифф», «посмотри изменения».
+  - Опираться на локальный репозиторий: ветка, `git diff`, открытые файлы.
+
+- **Что обязан проверить агент**
+  - **Архитектура и слои** (`architecture/boundaries.md`, `stack/ios-app-core.md`):
+    - View не вызывает API и не знает DTO.
+    - ViewModel не зависит от Data-реализаций напрямую.
+    - Domain не импортирует SwiftUI/UIKit.
+    - Data не тянет Presentation.
+  - **Зависимости и модули**:
+    - DI через протоколы; composition root в App.
+    - Нет deep-import между фичами.
+    - Файлы в правильных слоях (`Presentation`, `Domain`, `Data`, `DesignSystem`).
+  - **Swift и concurrency** (`stack/swift-conventions.md`):
+    - `@MainActor` на UI; нет необоснованных force unwrap.
+    - Typed errors; корректная отмена Task где нужно.
+  - **UI** (`ui-and-accessibility/swiftui.md`):
+    - Тонкие Views; Design System tokens; accessibility identifiers для тестируемых элементов.
+  - **Сеть и data** (`api-and-data/networking.md`):
+    - DTO отделены от domain; mappers покрывают edge cases.
+  - **Тесты**:
+    - Нетривиальная логика — unit (`testing/unit.md`); пользовательские потоки — UI (`testing/ui.md`).
+    - Указать, какие тесты добавить или обновить.
+
+- **Глубина и формат**
+  - Фокус на **диффе MR**, не на всём проекте.
+  - Следовать **Output contract** выше (`path:line`, severity, verdict).
+  - Без «больших рефакторингов» без запроса (`tooling-and-review/code-quality.md`).
+
+- **Ограничения**
+  - Не выдумывать метаданные CI/MR с хостинга без локальных данных.
+  - Boy scout rule в рамках MR.

package/presets/claude/ios-swift/rules/ui-and-accessibility/swiftui.md

@@ -20,6 +20,14 @@ paths:
 - View читает state и вызывает **явные методы** ViewModel (`submit()`, `load()`), а не мутирует внутреннее состояние напрямую.
 - Для навигации — coordinator/router или enum route в ViewModel, не размазанные `NavigationLink` с бизнес-условиями в View.
 
+# Lifecycle, ownership, hit testing
+
+- Prefer **`.task(id:)`** (auto-cancel) over `.onAppear { Task { … } }` for async loads tied to the view lifetime.
+- `@StateObject` / owned observable — **only where the view owns** the instance; prefer DI / parent-injected ViewModels over `@EnvironmentObject` sprawl.
+- Shared mutable services — `actor` (or existing isolation); default `[weak self]` in escaping closures that capture ViewModel/`self`.
+- Transparent stacks that should receive taps: `.contentShape(Rectangle())`.
+- High-frequency visual updates — prefer `TimelineView` (or repo pattern) over ad-hoc `Timer` in Views.
+
 # Структура файлов
 
 Для экрана фичи:

package/presets/claude/ios-swift/skills/README.md

@@ -9,5 +9,6 @@
 | `debug-investigation` | Runtime evidence before fix |
 | `ci-investigation` | iOS CI red |
 | `technical-retro` | Retro facilitation |
+| `write-adr` | Architecture Decision Record (context / decision / consequences) |
 
 Копируется в `.claude/skills/` при init.

package/presets/claude/ios-swift/skills/code-review/SKILL.md

@@ -18,11 +18,19 @@ Use the local repository as the source of truth: `git status`, `git diff`, task
 5. Review Swift concurrency and conventions (`stack/swift-conventions.md`): `@MainActor`, typed errors, force unwrap.
 6. Review SwiftUI thinness, Design System usage, accessibility identifiers (`ui-and-accessibility/swiftui.md`).
 7. Check tests: unit for mapper/domain/VM (`testing/unit.md`); XCUITest for user flows (`testing/ui.md`).
-8. Report validation gaps: `xcodebuild`, SwiftLint, or relevant tests not run (`tooling-and-review/xcode-tooling.md`, `tooling-and-review/post-change-build.md`).
+8. **Reuse lens:** flag new helpers that duplicate existing Feature/shared utilities; prefer public module APIs over copy-paste.
+9. **Efficiency lens:** unnecessary work on main thread, redundant network/persistence, hot-path bloat, missing cancellation/cleanup.
+10. Report validation gaps: `xcodebuild`, SwiftLint, or relevant tests not run (`tooling-and-review/xcode-tooling.md`, `tooling-and-review/post-change-build.md`).
 
 ## Output
 
-Lead with findings ordered by severity. Each finding should include the file, the problem, impact, and a concrete fix direction. If there are no findings, say so and call out remaining test or validation risk.
+Lead with findings ordered by severity. Required shape (also in `code-review`):
+
+- 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 `.claude/team/tasks/<slug>/review.md` and set `status.json` to `completed` or `changes_requested`.