@bonesofspring/ai-rules 0.1.39 → 0.1.40

package/README.md

@@ -26,7 +26,7 @@ Where things go on `init`:
 | agents | `.cursor/agents` | `.claude/agents` |
 | hooks | `.cursor/hooks`, `.cursor/hooks.json` | `.claude/hooks` |
 | skills | `.cursor/skills` | `.claude/skills` |
-| team artifacts template | `.cursor/team` | — |
+| team artifacts template | `.cursor/team` | `.claude/team` |
 | other | — | root `CLAUDE.md` → `.claude/CLAUDE.md` |
 
 The package currently ships the **`next`** preset (Next.js stack and whatever else is bundled there). Other preset names: see `presets/` in the [repo](https://github.com/bonesofspring/ai-rules). Name is a single path segment, no `/` or `..`.

package/bin/cli.js

@@ -41,6 +41,7 @@ Usage:
 Notes:
   init merges into existing target folders; same-named files are overwritten.
   Cursor init also copies agents/, hooks/, skills/, team/, and hooks.json when present.
+  Claude init also copies agents/, hooks/, skills/, team/, and CLAUDE.md when present.
   clean removes only paths installed by this CLI; other tool files are left intact.
   Flags may appear before or after the subcommand (e.g. --preset next init cursor).
 

package/package.json

@@ -1,12 +1,12 @@
 {
   "name": "@bonesofspring/ai-rules",
-  "version": "0.1.39",
+  "version": "0.1.40",
   "description": "Presets of Cursor and Claude rules/commands for Revy Ross personal use",
   "license": "MIT",
   "author": "Revy Ross",
   "repository": {
     "type": "git",
-    "url": "https://github.com/bonesofspring/ai-rules.git"
+    "url": "git+https://github.com/bonesofspring/ai-rules.git"
   },
   "keywords": [
     "cursor",
@@ -16,7 +16,7 @@
     "bonesofspring"
   ],
   "bin": {
-    "ai-rules": "./bin/cli.js"
+    "ai-rules": "bin/cli.js"
   },
   "files": [
     "README.md",

package/presets/claude/next/CLAUDE.md

@@ -15,6 +15,7 @@ See @rules/README.md for the full catalog and loading strategy.
 
 | Задача | Куда смотреть |
 |--------|----------------|
+| Любая новая work-задача | `commands/task.md` + `rules/tooling-and-review/agent-team-orchestrator.md` |
 | Новая фича end-to-end | `rules/architecture/feature-delivery-workflow.md` |
 | Стек и слои | `rules/stack/next-app-core.md` |
 | После любых правок кода | `rules/tooling-and-review/post-change-lint.md` |
@@ -27,4 +28,4 @@ See @rules/README.md for the full catalog and loading strategy.
 npx @bonesofspring/ai-rules init claude --preset next
 ```
 
-Копирует `CLAUDE.md`, `rules/**`, `commands/**`, `agents/**` и др. в `.claude/` целевого репозитория.
+Копирует `CLAUDE.md`, `rules/**`, `commands/**`, `agents/**`, `hooks/**`, `team/**` и др. в `.claude/` целевого репозитория.

package/presets/claude/next/README.md

@@ -10,5 +10,15 @@
 | `skills/**` | `.claude/skills/**` (если папка есть) |
 | `agents/**` | `.claude/agents/**` (если папка есть) |
 | `hooks/**` | `.claude/hooks/**` (если папка есть) |
+| `team/**` | `.claude/team/**` (шаблон артефактов agent team) |
 
 Файлы правил — обычный Markdown с опциональным YAML frontmatter (`paths:` для скоупа по путям), не формат `.mdc` Cursor.
+
+## Agent team workflow
+
+Cursor-пресет и Claude-пресет используют один и тот же смысловой workflow:
+
+- `/task <desc>` → `task-router` → `pipeline.json`;
+- `/task-continue <slug>` → продолжение после human gate;
+- `.claude/agents/*.md` → роли analyst / architect / debugger / developer / reviewer / QA;
+- `.claude/team/tasks/<slug>/` → handoff-артефакты (`brief.md`, `decomposition.md`, `review.md`, `status.json`).

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

@@ -3,3 +3,18 @@
 Специализированные подагенты (personas) — по [документации Claude Code](https://code.claude.com/docs), если используете их в проекте.
 
 Папка копируется в `.claude/agents/` при `ai-rules init claude --preset next`, если существует.
+
+| Agent | Может менять production-код | Назначение |
+|-------|-----------------------------|------------|
+| `task-router` | no | Классифицирует intent и пишет `pipeline.json` |
+| `task-analyst` | no | DoR, acceptance criteria, decomposition |
+| `solution-architect` | no | Cross-layer design, spike, architecture.md |
+| `debugger` | yes, минимально | Root cause, debug-report.md, простые fixes |
+| `feature-developer` | yes | Реализация по decomposition |
+| `code-reviewer` | no | Review artifact, status update |
+| `qa-tester` | tests only | Unit/e2e тесты |
+| `playwright-test-planner` | no | Планирование e2e сценариев |
+| `playwright-test-generator` | tests only | Генерация Playwright specs |
+| `playwright-test-healer` | tests only | Исправление flaky/failing e2e |
+
+Оркестрация: `/task`, `/task-continue`, `rules/tooling-and-review/agent-team-orchestrator.md`, артефакты `.claude/team/**`.

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

@@ -4,6 +4,10 @@
 
 | Файл | Назначение |
 |------|------------|
-| `technical-retro.md` | Фасилитация технического ретро (не дублируется в `rules/` — экономия контекста) |
+| `task.md` | Единая точка входа: router → dynamic pipeline → первый agent |
+| `task-continue.md` | Продолжение после human gate / паузы по `pipeline.json` |
+| `feature-start.md` | Legacy analyst-only старт без router |
+| `feature-continue.md` | Legacy продолжение; для `pipeline.json` делегирует смысл `/task-continue` |
+| `technical-retro.md` | Фасилитация технического ретро, включая блок «Работа агентов» |
 
-См. также корневой `README.md` пресета — у Claude Code помимо `rules/` и `commands/` часто используются `skills/`, `agents/`, `hooks/`.
+См. также корневой `README.md` пресета — у Claude Code помимо `rules/` и `commands/` часто используются `skills/`, `agents/`, `hooks/`, `team/`.

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

@@ -3,3 +3,9 @@
 Событийные хуки автоматизации Claude Code — по документации продукта, если включены в репозитории.
 
 Папка копируется в `.claude/hooks/` при `ai-rules init claude --preset next`, если существует.
+
+| Файл | Назначение |
+|------|------------|
+| `chain-team-phases.sh` | Optional chaining helper для agent team pipeline после завершения subagent |
+
+Hook-скрипт сам по себе не включает автоматизацию: подключайте его в конфигурации Claude Code hooks целевого проекта, если хотите auto-chain. Без hook workflow продолжает родительский агент по `rules/tooling-and-review/agent-team-orchestrator.md`.

package/presets/claude/next/hooks/chain-team-phases.sh

@@ -0,0 +1,123 @@
+#!/usr/bin/env bash
+# Chains agent team phases after subagent completion when wired into Claude Code hooks.
+# Reads stdin JSON from a subagent-stop style hook; writes followup_message to stdout.
+
+set -euo pipefail
+
+input=$(cat)
+
+INPUT="$input" node <<'NODE'
+const fs = require('fs');
+const path = require('path');
+
+const inputRaw = process.env.INPUT || '';
+let hookInput;
+try {
+  hookInput = JSON.parse(inputRaw);
+} catch {
+  process.stdout.write('{}');
+  process.exit(0);
+}
+
+if (hookInput.status !== 'completed') {
+  process.stdout.write('{}');
+  process.exit(0);
+}
+
+const teamRoot = path.join(process.cwd(), '.claude', 'team');
+const activePath = path.join(teamRoot, 'active-task.json');
+
+if (!fs.existsSync(activePath)) {
+  process.stdout.write('{}');
+  process.exit(0);
+}
+
+let active;
+try {
+  active = JSON.parse(fs.readFileSync(activePath, 'utf8'));
+} catch {
+  process.stdout.write('{}');
+  process.exit(0);
+}
+
+const slug = active.slug;
+if (!slug) {
+  process.stdout.write('{}');
+  process.exit(0);
+}
+
+const statusPath = path.join(teamRoot, 'tasks', slug, 'status.json');
+if (!fs.existsSync(statusPath)) {
+  process.stdout.write('{}');
+  process.exit(0);
+}
+
+let status;
+try {
+  status = JSON.parse(fs.readFileSync(statusPath, 'utf8'));
+} catch {
+  process.stdout.write('{}');
+  process.exit(0);
+}
+
+const now = new Date().toISOString();
+
+function writeStatus(next) {
+  fs.writeFileSync(statusPath, JSON.stringify({ ...status, ...next, slug, updatedAt: now }, null, 2) + '\n');
+}
+
+function out(message) {
+  process.stdout.write(JSON.stringify({ followup_message: message }));
+}
+
+// Human gate: do not auto-chain during analysis or explicit approval waits.
+if (status.phase === 'analysis' || status.state === 'awaiting_approval') {
+  process.stdout.write('{}');
+  process.exit(0);
+}
+
+// Legacy mode without pipeline.json.
+if (status.phase === 'development' && status.state === 'completed') {
+  writeStatus({ phase: 'review', state: 'in_progress' });
+  out(
+    `Subagent development phase finished for slug "${slug}". ` +
+      `Invoke the code-reviewer subagent. ` +
+      `Read .claude/team/tasks/${slug}/brief.md and decomposition.md. ` +
+      `Review git diff against acceptance criteria per rules/tooling-and-review/code-review-mr.md. ` +
+      `Update status.json when done.`,
+  );
+  process.exit(0);
+}
+
+if (status.phase === 'review' && status.state === 'completed') {
+  writeStatus({ phase: 'testing', state: 'in_progress' });
+  out(
+    `Code review completed for slug "${slug}". ` +
+      `Invoke the qa-tester subagent. ` +
+      `Add or update unit and e2e tests per brief acceptance criteria. ` +
+      `Follow rules/testing/playwright-agents.md and rules/testing/tests-e2e-structure.md. ` +
+      `Update status.json when done.`,
+  );
+  process.exit(0);
+}
+
+if (status.phase === 'review' && status.state === 'changes_requested') {
+  out(
+    `Review requested changes for slug "${slug}". ` +
+      `Invoke feature-developer with the review fix list, then re-run code-reviewer.`,
+  );
+  process.exit(0);
+}
+
+if (status.phase === 'testing' && status.state === 'completed') {
+  writeStatus({ phase: 'done', state: 'completed' });
+  out(
+    `QA phase completed for slug "${slug}". ` +
+      `Suggest running /technical-retro with this slug. ` +
+      `Read team artifacts under .claude/team/tasks/${slug}/ and git diff for the retro.`,
+  );
+  process.exit(0);
+}
+
+process.stdout.write('{}');
+NODE

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

@@ -17,13 +17,15 @@ On-demand сценарии вне постоянного контекста —
 
 ## С чего начать новую фичу
 
-**`architecture/feature-delivery-workflow.md`** — сквозной чеклист и матрица «зона репозитория → правило».
+Для обычной задачи — **`commands/task.md`**: router → `pipeline.json` → нужные agents. Для ручной реализации без team workflow — **`architecture/feature-delivery-workflow.md`**.
 
 ## Каталог правил
 
 | Файл | Назначение |
 |------|------------|
 | `tooling-and-review/package-manager.md` | Менеджер пакетов перед install/run |
+| `tooling-and-review/agent-team-intake.md` | Когда использовать `/task` и task-router |
+| `tooling-and-review/agent-team-orchestrator.md` | Dynamic pipeline, subagents, gates, handoff artifacts |
 | `architecture/feature-delivery-workflow.md` | Сквозной порядок работ по фиче, mermaid, матрица слой → правило |
 | `stack/next-app-core.md` | Стек, слои, порты‑адаптеры, работа агента |
 | `tooling-and-review/post-change-lint.md` | **Обязательный** ESLint + Stylelint после изменений |

package/presets/claude/next/rules/tooling-and-review/README.md

@@ -5,6 +5,8 @@ Lint, package manager, качество кода, code review MR.
 | Файл | Содержание |
 |------|------------|
 | `package-manager.md` | yarn / pnpm / npm / bun — определить по lockfile |
+| `agent-team-intake.md` | Когда предлагать `/task` и router pipeline |
+| `agent-team-orchestrator.md` | Dynamic agent team workflow: router, pipeline, gates |
 | `post-change-lint.md` | Обязательный `lint:js` + `lint:css` после изменений |
 | `code-quality.md` | Boy scout rule, рефакторинг, ESLint |
 | `code-review-mr.md` | Чеклист ревью merge request |