@torus-engineering/tas-kit 1.10.0 → 1.12.0

package/README.md

@@ -1,45 +1,45 @@
 # TAS Kit
 
-**Torus Agentic SDLC Kit** — Bộ công cụ AI-assisted SDLC cho Claude Code, được thiết kế theo triết lý Spec-Driven Development và Context-Aware Architecture.
+**Torus Agentic SDLC Kit** — AI-assisted SDLC toolkit, designed following Spec-Driven Development philosophy and Context-Aware Architecture.
 
 ---
 
-## Mục đích
+## Purpose
 
-TAS Kit giúp các team phát triển phần mềm:
-- **Chuẩn hóa quy trình SDLC** với các artifacts rõ ràng (PRD, SAD, ADR, Epic, Feature, Story)
-- **Tối ưu hóa token usage** thông qua Context Layer — đầu tư ở phase thiết kế, tiết kiệm ở phase code
-- **Hỗ trợ nhiều mô hình làm việc**: từ solo developer đến full team với các role PE/SE/DSE
-- **Tự động hóa** với Hybrid và Autonomous agents hoạt động 24/7
+TAS Kit helps software development teams:
+- **Standardize SDLC process** with clear artifacts (PRD, SAD, ADR, Epic, Feature, Story)
+- **Optimize token usage** through Context Layer — invest in design phase, save in code phase
+- **Support multiple workflows**: from solo developer to full team with PE/SE/DSE roles
+- **Automate** with Hybrid and Autonomous agents operating 24/7
 
 ---
 
-## Triết lý thiết kế
+## Design Philosophy
 
 ### Spec-Driven Development
-Tất cả code bắt đầu từ spec. Story.md là "context digest" đã hấp thụ toàn bộ thông tin từ PRD, SAD, ADR, Feature. Mỗi session implement chỉ cần đọc CLAUDE.md + Story.md — không cần load lại toàn bộ artifacts.
+All code starts from spec. Story.md is "context digest" that has absorbed all information from PRD, SAD, ADR, Feature. Each implementation session only needs to read CLAUDE.md + Story.md — no need to reload all artifacts.
 
 ### Human-Led, AI-Powered
-- **Con người chủ đạo**: Quyết định architecture, approve design, review code
-- **AI hỗ trợ**: Tạo draft, implement theo spec, chạy tests, phát hiện bugs
+- **Human-led**: Architecture decisions, design approval, code review
+- **AI-assisted**: Draft creation, implementation per spec, test execution, bug detection
 
 ### Dual Mode Operation
-- **Hybrid Mode**: Developer và AI làm việc cùng nhau trong session
-- **Autonomous Mode**: AI tự động develop 24/7, report kết quả khi xong
+- **Hybrid Mode**: Developer and AI work together in session
+- **Autonomous Mode**: AI automatically develops 24/7, reports results when done
 
 ### Lightweight & Context-Aware
-- Mỗi skill < 3KB, tổng kit < 50KB
-- Context Layer: đầu tư token ở phase thiết kế để tiết kiệm gấp nhiều lần ở phase code
-- Khi code: PHẢI bắt đầu session mới (không tận dụng session cũ đã load PRD/SAD vì sẽ phình Window Context)
+- Each skill < 3KB, total kit < 50KB
+- Context Layer: invest tokens in design phase to save many times in code phase
+- When coding: MUST start new session (don't reuse old session that loaded PRD/SAD to avoid Window Context bloat)
 
 ### Role-Based & Template-Driven
-- 3 role thực tế của Torus: PE (Product Engineer), SE (Software Engineer), DSE (DevOps Engineer)
-- Templates riêng cho PRD, SAD, ADR, Epic, Feature, Story theo chuẩn Torus
-- Flow configurable qua file `tas.yaml`
+- 3 actual roles at Torus: PE (Product Engineer), SE (Software Engineer), DSE (DevOps Engineer)
+- Separate templates for PRD, SAD, ADR, Epic, Feature, Story per Torus standards
+- Flow configurable via `tas.yaml` file
 
 ### Azure DevOps Compatible
-- Output Markdown tương thích Azure DevOps Wiki
-- Mermaid diagrams không dùng ký tự `()` (tránh conflict với ADO syntax)
+- Markdown output compatible with Azure DevOps Wiki
+- Mermaid diagrams don't use `()` characters (avoid conflict with ADO syntax)
 
 ---
 
@@ -57,7 +57,7 @@ graph TD
         PRD_ARTIFACT -->|Software Engineer| ADR
         SAD --> SAD_ARTIFACT["SAD.md"]
         ADR --> ADR_ARTIFACT["ADR-XXX.md"]
-        PRD_ARTIFACT -->|Software Engineer| DESIGN
+        PRD_ARTIFACT -->|Product Engineer| DESIGN
         DESIGN --> DESIGN_ARTIFACT["design-spec.md"]
     end
 
@@ -65,17 +65,14 @@ graph TD
         EPIC["/tas-epic"]
         FEATURE["/tas-feature"]
         STORY["/tas-story"]
-        FUNctest["/tas-functest"]
-        PRD_ARTIFACT -->|Software Engineer| EPIC
+        PRD_ARTIFACT -->|Product Engineer| EPIC
         EPIC --> EPIC_ARTIFACT["Epic-XXX/"]
-        EPIC_ARTIFACT -->|Software Engineer| FEATURE
+        EPIC_ARTIFACT -->|Product Engineer| FEATURE
         FEATURE --> FEATURE_ARTIFACT["Feature-XXX/"]
         SAD_ARTIFACT --> FEATURE
         DESIGN_ARTIFACT --> FEATURE
-        FEATURE_ARTIFACT -->|Software Engineer| STORY
+        FEATURE_ARTIFACT -->|Product Engineer| STORY
         STORY --> STORY_ARTIFACT["Story-XXX.md"]
-        FEATURE_ARTIFACT -->|Software Engineer| FUNctest
-        FUNctest --> FUNCTEST_ARTIFACT["Functional-Testing-Spec-FeatureXXX.md"]
     end
 
     subgraph Phase3["Phase 3: Implementation"]
@@ -90,9 +87,10 @@ graph TD
         E2E_MOBILE["/tas-e2e-mobile"]
         BUG["/tas-bug"]
         REVIEW["/tas-review"]
+        FUNctest["/tas-functest"]
         STORY_ARTIFACT -->|Software Engineer| PLAN
-        PLAN --> PLAN_ARTIFACT["Plan.md"]
-        PLAN_ARTIFACT -->|AI Agent| DEV
+        PLAN --> STORY_ARTIFACT
+        STORY_ARTIFACT -->|AI Agent| DEV
         DEV --> CODE["Source Code"]
         CODE -->|Bug found| FIX
         FIX --> CODE
@@ -112,6 +110,8 @@ graph TD
         E2E_MOBILE --> E2E_MOBILE_ARTIFACT["E2E Mobile Test Script"]
         CODE -->|Bug found| BUG
         BUG --> BUG_ARTIFACT["Bug Report"]
+        FEATURE_ARTIFACT -->|Product Engineer| FUNctest
+        FUNctest --> FUNCTEST_ARTIFACT["Functional-Testing-Spec-FeatureXXX.md"]
     end
 
     subgraph Phase4["Phase 4: Quality & Deploy"]
@@ -152,29 +152,29 @@ graph TD
 | **Phase 3: Implementation** | PE, SE, AI | `/tas-plan`, `/tas-dev`, `/tas-fix`, `/tas-apitest`, `/tas-e2e`, `/tas-functest-web`, `/tas-functest-mobile`, `/tas-e2e-web`, `/tas-e2e-mobile`, `/tas-bug`, `/tas-review` | Plan.md, Source Code, API Testing Script, E2E Test Scenarios, Functional Test Scripts, Bug Report, Code Review Report |
 | **Phase 4: Quality & Deploy** | PE, DSE | `/tas-security`, Pipeline/CLI | Security Report, Test Report, Production |
 
-### Chi tiết từng Phase
+### Phase Details
 
 #### Phase 1: Discovery
-- **PE**: Tạo PRD với `/tas-prd`
-- **SE**: Tạo SAD với `/tas-sad`, tạo ADR với `/tas-adr`, tạo design-spec với `/tas-design`
+- **PE**: Create PRD with `/tas-prd`, create design-spec with `/tas-design`
+- **SE**: Create SAD with `/tas-sad`, create ADR with `/tas-adr`
 
 #### Phase 2: Planning
-- **SE**: Breakdown Epic với `/tas-epic`, Feature với `/tas-feature`, Story với `/tas-story`
-- **SE**: Tạo Functional Testing Spec với `/tas-functest`
+- **PE**: Breakdown Epic with `/tas-epic`, Feature with `/tas-feature`, Story with `/tas-story`
 
 #### Phase 3: Implementation
-- **SE**: Tạo Plan với `/tas-plan`, Implement với `/tas-dev`, Fix bug với `/tas-fix`
-- **SE**: Review code với `/tas-review`
-- **SE**: Tạo API Testing Script với `/tas-apitest`
-- **PE**: Tạo E2E Test Scenarios với `/tas-e2e`
-- **SE**: Tạo Functional Test Scripts với `/tas-functest-web`, `/tas-functest-mobile`
-- **PE**: Tạo E2E Test Scripts với `/tas-e2e-web`, `/tas-e2e-mobile`
-- **PE**: Tạo Bug Report với `/tas-bug`
+- **SE**: Create Plan with `/tas-plan`, Implement with `/tas-dev`, Fix bug with `/tas-fix`
+- **SE**: Review code with `/tas-review`
+- **PE**: Create Functional Testing Spec with `/tas-functest`
+- **SE**: Create API Testing Script with `/tas-apitest`
+- **PE**: Create E2E Test Scenarios with `/tas-e2e`
+- **SE**: Create Functional Test Scripts with `/tas-functest-web`, `/tas-functest-mobile`
+- **PE**: Create E2E Test Scripts with `/tas-e2e-web`, `/tas-e2e-mobile`
+- **PE**: Create Bug Report with `/tas-bug`
 
 #### Phase 4: Quality & Deploy
-- **DSE**: Security Review với `/tas-security`
-- **PE**: Chạy Pipeline/CLI để execute Automation Tests
-- **DSE**: Deploy lên Production
+- **DSE**: Security Review with `/tas-security`
+- **PE**: Run Pipeline/CLI to execute Automation Tests
+- **DSE**: Deploy to Production
 
 ---
 
@@ -193,13 +193,13 @@ npx @torus-engineering/tas-kit install --yes    # skip confirmation prompts
 
 ---
 
-## Thiết lập quan trọng
+## Important Setup
 
 ### CLAUDE.md
 
-File cấu hình quan trọng nhất — Claude đọc file này đầu tiên trong mọi session.
+Most important configuration file — Claude reads this file first in every session.
 
-**Cấu trúc:**
+**Structure:**
 ```markdown
 # Project Name
 
@@ -217,9 +217,9 @@ File cấu hình quan trọng nhất — Claude đọc file này đầu tiên tr
 
 ### tas.yaml
 
-Điều khiển flow của TAS Kit theo từng dự án.
+Controls TAS Kit flow per project.
 
-**Cấu trúc:**
+**Structure:**
 ```yaml
 project:
   name: "My Project"
@@ -238,9 +238,9 @@ flow:
 
 ### project-status.yaml
 
-Index tổng hợp trạng thái dự án — được cập nhật tự động sau mỗi thay đổi artifact.
+Project status index — automatically updated after each artifact change.
 
-**Cấu trúc:**
+**Structure:**
 ```yaml
 last_updated: 2025-01-15
 
@@ -262,7 +262,7 @@ epics:
 
 ### .env
 
-Environment variables cho Azure DevOps integration.
+Environment variables for Azure DevOps integration.
 
 ```bash
 AZURE_DEVOPS_PAT=your_pat_here
@@ -281,7 +281,7 @@ AZURE_DEVOPS_PROJECT=your_project
   agents/     29 specialized subagents
 .tas/
   templates/  Markdown templates (PRD, SAD, ADR, Epic, Feature, Story...)
-  checklists/ Code review, story done, security checklists
+  rules/      Coding standards + workflow rules (code-review, story-done, security...)
   tools/      ADO integration script (tas-ado.py)
 CLAUDE.md     Project context template (edit this for your project)
 tas.yaml      Flow configuration template (edit this for your project)
@@ -292,31 +292,31 @@ tas.yaml      Flow configuration template (edit this for your project)
 
 ## Commands Reference
 
-| Command | Mô tả | Role |
-|---------|-------|------|
-| `/tas-init` | Khởi tạo TAS cho dự án mới | All |
-| `/tas-status` | Hiển thị trạng thái hiện tại của dự án | All |
-| `/tas-prd` | Tạo/cập nhật Product Requirements Document | PE |
-| `/tas-sad` | Tạo/cập nhật Solution Architecture Document | SE |
-| `/tas-adr` | Tạo Architecture Decision Record | SE |
-| `/tas-design` | Tạo Design Specification | SE |
-| `/tas-epic` | Tạo/cập nhật Epic | SE |
-| `/tas-feature` | Tạo/cập nhật Feature | SE |
-| `/tas-story` | Tạo/cập nhật Story | SE |
-| `/tas-functest` | Tạo Functional Testing Specification | SE |
+| Command | Description | Role |
+|---------|-------------|------|
+| `/tas-init` | Initialize TAS for new project | All |
+| `/tas-status` | Display current project status | All |
+| `/tas-prd` | Create/update Product Requirements Document | PE |
+| `/tas-sad` | Create/update Solution Architecture Document | SE |
+| `/tas-adr` | Create Architecture Decision Record | SE |
+| `/tas-design` | Create Design Specification | SE |
+| `/tas-epic` | Create/update Epic | SE |
+| `/tas-feature` | Create/update Feature | SE |
+| `/tas-story` | Create/update Story | SE |
+| `/tas-functest` | Create Functional Testing Specification | SE |
 | `/tas-spec` | Lightweight spec (solo / prototype) | SE |
-| `/tas-plan` | Lập kế hoạch kỹ thuật implementation | SE |
+| `/tas-plan` | Create technical implementation plan | SE |
 | `/tas-dev` | Implement story (agentic) | AI |
-| `/tas-fix` | Quick fix không cần full story flow | SE |
-| `/tas-apitest` | Tạo API Testing Script tự động | SE |
-| `/tas-e2e` | Tạo E2E Test Scenarios | PE |
-| `/tas-functest-web` | Tạo Functional Test Script cho Web | SE |
-| `/tas-functest-mobile` | Tạo Functional Test Script cho Mobile | SE |
-| `/tas-e2e-web` | Tạo E2E Test Script cho Web | PE |
-| `/tas-e2e-mobile` | Tạo E2E Test Script cho Mobile | PE |
-| `/tas-bug` | Tạo Bug Report | PE |
-| `/tas-review` | Code Review với checklist | SE |
-| `/tas-brainstorm` | Brainstorm giải pháp | All |
+| `/tas-fix` | Quick fix without full story flow | SE |
+| `/tas-apitest` | Create API Testing Script automatically | SE |
+| `/tas-e2e` | Create E2E Test Scenarios | PE |
+| `/tas-functest-web` | Create Functional Test Script for Web | SE |
+| `/tas-functest-mobile` | Create Functional Test Script for Mobile | SE |
+| `/tas-e2e-web` | Create E2E Test Script for Web | PE |
+| `/tas-e2e-mobile` | Create E2E Test Script for Mobile | PE |
+| `/tas-bug` | Create Bug Report | PE |
+| `/tas-review` | Code Review with checklist | SE |
+| `/tas-brainstorm` | Brainstorm solutions | All |
 | `/tas-security` | Security Review | DSE |
 | `/ado-*` | Azure DevOps integration | All |
 
@@ -331,4 +331,4 @@ tas.yaml      Flow configuration template (edit this for your project)
 
 ## Documentation
 
-Xem `.tas/README.md` sau khi install để có tài liệu chi tiết.
+See `.tas/README.md` after install for detailed documentation.

package/bin/cli.js

@@ -1,73 +1,91 @@
-#!/usr/bin/env node
-import { install, update } from '../lib/install.js';
-
-const args = process.argv.slice(2);
-const command = args[0];
-
-function printHelp() {
-  console.log(`
-tas-kit — Torus Agentic SDLC Kit installer
-
-Usage:
-  npx @torus-engineering/tas-kit <command> [options]
-
-Commands:
-  install   Copy TAS Kit files into a project (first-time setup)
-  update    Overwrite .claude/ and .tas/ with the latest kit version
-            (preserves CLAUDE.md, tas.yaml, .env.example)
-
-Options:
-  --directory <path>          Target directory (default: current working directory)
-  --yes, -y                   Skip confirmation prompts (uses native hook by default)
-  --security-hook <mode>      Pre-commit hook mode: husky | native | none
-                              Default: interactive prompt on install,
-                                       no-op on update unless specified.
-  --help, -h                  Show this help message
-
-Examples:
-  npx @torus-engineering/tas-kit install
-  npx @torus-engineering/tas-kit install --directory /path/to/my-project
-  npx @torus-engineering/tas-kit install --yes --security-hook=husky
-  npx @torus-engineering/tas-kit install --security-hook=none
-  npx @torus-engineering/tas-kit update
-  npx @torus-engineering/tas-kit update --yes --security-hook=native
-  `.trim());
-}
-
-if (!command || command === '--help' || command === '-h') {
-  printHelp();
-  process.exit(0);
-}
-
-if (command !== 'install' && command !== 'update') {
-  console.error(`Unknown command: "${command}"\n`);
-  printHelp();
-  process.exit(1);
-}
-
-const opts = { directory: process.cwd(), yes: false, securityHook: null };
-
-for (let i = 1; i < args.length; i++) {
-  const a = args[i];
-  if ((a === '--directory' || a === '-d') && args[i + 1]) {
-    opts.directory = args[++i];
-  } else if (a === '--yes' || a === '-y') {
-    opts.yes = true;
-  } else if (a.startsWith('--security-hook=')) {
-    opts.securityHook = a.slice('--security-hook='.length).toLowerCase();
-  } else if (a === '--security-hook' && args[i + 1]) {
-    opts.securityHook = args[++i].toLowerCase();
-  }
-}
-
-if (opts.securityHook && !['husky', 'native', 'none'].includes(opts.securityHook)) {
-  console.error(`Invalid --security-hook value: "${opts.securityHook}"`);
-  console.error(`Expected one of: husky, native, none`);
-  process.exit(1);
-}
-
-const runner = command === 'update' ? update : install;
-runner(opts).catch((err) => {
-  console.error(`\n${command === 'update' ? 'Update' : 'Install'} failed:`, err.message);
-  process.exit(1);
-});
+#!/usr/bin/env node
+import { install, update } from '../lib/install.js';
+
+const args = process.argv.slice(2);
+const command = args[0];
+
+function printHelp() {
+  console.log(`
+tas-kit — Torus Agentic SDLC Kit installer
+
+Usage:
+  npx @torus-engineering/tas-kit <command> [options]
+
+Commands:
+  install   Copy TAS Kit files into a project (first-time setup)
+  update    Overwrite .tas/ and platform directories with the latest kit version
+            (preserves CLAUDE.md, tas.yaml, .env.example)
+
+Options:
+  --directory <path>          Target directory (default: current working directory)
+  --yes, -y                   Skip confirmation prompts
+  --platform <id,...>         Platform(s) to install: claude-code, cursor, codex, antigravity
+                              Comma-separated for multiple (default: interactive prompt)
+  --security-hook <mode>      Pre-commit hook mode: husky | native | none
+  --help, -h                  Show this help message
+
+Platforms:
+  claude-code    Claude Code CLI (default)
+  cursor         Cursor IDE
+  codex          OpenAI Codex CLI
+  antigravity    Google Antigravity IDE
+
+Examples:
+  npx @torus-engineering/tas-kit install
+  npx @torus-engineering/tas-kit install --platform claude-code,cursor
+  npx @torus-engineering/tas-kit install --platform codex --yes
+  npx @torus-engineering/tas-kit install --yes --security-hook=husky
+  npx @torus-engineering/tas-kit update
+  npx @torus-engineering/tas-kit update --platform cursor
+  `.trim());
+}
+
+if (!command || command === '--help' || command === '-h') {
+  printHelp();
+  process.exit(0);
+}
+
+if (command !== 'install' && command !== 'update') {
+  console.error(`Unknown command: "${command}"\n`);
+  printHelp();
+  process.exit(1);
+}
+
+const VALID_PLATFORMS = ['claude-code', 'cursor', 'codex', 'antigravity'];
+const opts = { directory: process.cwd(), yes: false, securityHook: null, platforms: [] };
+
+for (let i = 1; i < args.length; i++) {
+  const a = args[i];
+  if ((a === '--directory' || a === '-d') && args[i + 1]) {
+    opts.directory = args[++i];
+  } else if (a === '--yes' || a === '-y') {
+    opts.yes = true;
+  } else if (a.startsWith('--security-hook=')) {
+    opts.securityHook = a.slice('--security-hook='.length).toLowerCase();
+  } else if (a === '--security-hook' && args[i + 1]) {
+    opts.securityHook = args[++i].toLowerCase();
+  } else if (a.startsWith('--platform=')) {
+    opts.platforms = a.slice('--platform='.length).split(',').map(s => s.trim().toLowerCase());
+  } else if (a === '--platform' && args[i + 1]) {
+    opts.platforms = args[++i].split(',').map(s => s.trim().toLowerCase());
+  }
+}
+
+if (opts.securityHook && !['husky', 'native', 'none'].includes(opts.securityHook)) {
+  console.error(`Invalid --security-hook value: "${opts.securityHook}"`);
+  console.error(`Expected one of: husky, native, none`);
+  process.exit(1);
+}
+
+const invalidPlatforms = opts.platforms.filter(p => !VALID_PLATFORMS.includes(p));
+if (invalidPlatforms.length > 0) {
+  console.error(`Invalid platform(s): ${invalidPlatforms.join(', ')}`);
+  console.error(`Valid platforms: ${VALID_PLATFORMS.join(', ')}`);
+  process.exit(1);
+}
+
+const runner = command === 'update' ? update : install;
+runner(opts).catch((err) => {
+  console.error(`\n${command === 'update' ? 'Update' : 'Install'} failed:`, err.message);
+  process.exit(1);
+});

package/lib/adapters/antigravity.js

@@ -0,0 +1,137 @@
+import fs from 'node:fs/promises';
+import path from 'node:path';
+import {
+  exists, listFilesRecursive,
+  parseFrontmatter, extractDescription, extractCommandName, extractH1, yamlValue,
+} from './utils.js';
+
+export const PLATFORM_ID = 'antigravity';
+export const PLATFORM_LABEL = 'Antigravity';
+
+export async function install({ tasDir, target }) {
+  const agentsDir = path.join(target, '.agents');
+  await fs.mkdir(agentsDir, { recursive: true });
+  await fs.mkdir(path.join(agentsDir, 'rules'), { recursive: true });
+  await fs.mkdir(path.join(agentsDir, 'skills'), { recursive: true });
+
+  // commands → .agents/<name>.md (Antigravity workflows)
+  const commandsDir = path.join(tasDir, 'commands');
+  if (await exists(commandsDir)) {
+    const files = await listFilesRecursive(commandsDir, '.md');
+    for (const file of files) {
+      await commandToWorkflow(file, agentsDir);
+    }
+    console.log('  [ok] .agents/ (workflows from commands)');
+  }
+
+  // agents → .agents/skills/<name>/SKILL.md
+  const agentsSrc = path.join(tasDir, 'agents');
+  if (await exists(agentsSrc)) {
+    const files = await listFilesRecursive(agentsSrc, '.md');
+    for (const file of files) {
+      await agentToSkill(file, path.join(agentsDir, 'skills'));
+    }
+    console.log('  [ok] .agents/skills/ (agents)');
+  }
+
+  // kit skills → .agents/skills/ (copy as-is, same open standard)
+  const skillsSrc = path.join(tasDir, 'skills');
+  if (await exists(skillsSrc)) {
+    const skillDirs = await fs.readdir(skillsSrc, { withFileTypes: true });
+    for (const d of skillDirs.filter(e => e.isDirectory())) {
+      await fs.cp(
+        path.join(skillsSrc, d.name),
+        path.join(agentsDir, 'skills', d.name),
+        { recursive: true }
+      );
+    }
+    console.log('  [ok] .agents/skills/ (skills)');
+  }
+
+  // rules → .agents/rules/<stack>-<topic>.md
+  const rulesDir = path.join(tasDir, 'rules');
+  if (await exists(rulesDir)) {
+    await convertRules(rulesDir, path.join(agentsDir, 'rules'));
+    console.log('  [ok] .agents/rules/ (from .tas/rules/)');
+  }
+
+  // lean AGENTS.md
+  await generateAgentsMd(agentsDir, target);
+  console.log('  [ok] AGENTS.md');
+}
+
+async function commandToWorkflow(file, agentsDir) {
+  const content = await fs.readFile(file, 'utf8');
+  const name = extractCommandName(content) || path.basename(file, '.md');
+  // Antigravity workflows: same .md format, stored in .agents/<name>.md
+  await fs.writeFile(path.join(agentsDir, `${name}.md`), content);
+}
+
+async function agentToSkill(file, skillsDir) {
+  const content = await fs.readFile(file, 'utf8');
+  const { frontmatter, body } = parseFrontmatter(content);
+  const name = frontmatter.name || path.basename(file, '.md');
+  const description = frontmatter.description || extractDescription(body, name);
+  const skillDir = path.join(skillsDir, name);
+  await fs.mkdir(skillDir, { recursive: true });
+  // Antigravity SKILL.md: same open standard, drop allowed-tools (platform manages tool access)
+  const skillMd = `---\nname: ${yamlValue(name)}\ndescription: ${yamlValue(description)}\n---\n\n${body}`;
+  await fs.writeFile(path.join(skillDir, 'SKILL.md'), skillMd);
+}
+
+async function convertRules(rulesDir, outDir) {
+  const entries = await fs.readdir(rulesDir, { withFileTypes: true });
+  for (const entry of entries) {
+    if (entry.isDirectory()) {
+      const stack = entry.name;
+      const files = await listFilesRecursive(path.join(rulesDir, stack), '.md');
+      for (const f of files) {
+        const topic = path.basename(f, '.md');
+        await fs.copyFile(f, path.join(outDir, `${stack}-${topic}.md`));
+      }
+    } else if (entry.name.endsWith('.md')) {
+      await fs.copyFile(path.join(rulesDir, entry.name), path.join(outDir, entry.name));
+    }
+  }
+}
+
+async function generateAgentsMd(agentsDir, target) {
+  const workflowFiles = (await fs.readdir(agentsDir, { withFileTypes: true }))
+    .filter(e => e.isFile() && e.name.endsWith('.md'))
+    .map(e => `- \`/${e.name.replace('.md', '')}\``);
+
+  const skillDirs = await fs.readdir(path.join(agentsDir, 'skills'), { withFileTypes: true });
+  const skillLines = [];
+  for (const d of skillDirs.filter(e => e.isDirectory())) {
+    const skillFile = path.join(agentsDir, 'skills', d.name, 'SKILL.md');
+    if (!(await exists(skillFile))) continue;
+    const content = await fs.readFile(skillFile, 'utf8');
+    const { frontmatter } = parseFrontmatter(content);
+    if (frontmatter.name) {
+      skillLines.push(`- **${frontmatter.name}**: ${frontmatter.description || ''}`);
+    }
+  }
+
+  const agentsMd = [
+    `# Project Instructions`,
+    ``,
+    `This project uses TAS Kit — Torus Agentic SDLC Kit.`,
+    ``,
+    `## Workflows`,
+    ``,
+    `Invoke via \`/workflow-name\` in chat:`,
+    ``,
+    ...workflowFiles,
+    ``,
+    `## Skills`,
+    ``,
+    ...skillLines,
+    ``,
+    `## Project Configuration`,
+    ``,
+    `See \`tas.yaml\` for project settings and project instructions file for detailed conventions.`,
+    ``,
+  ].join('\n');
+
+  await fs.writeFile(path.join(target, 'AGENTS.md'), agentsMd);
+}

package/lib/adapters/claude-code.js

@@ -0,0 +1,35 @@
+import fs from 'node:fs/promises';
+import path from 'node:path';
+import { copyDir, exists } from './utils.js';
+
+export const PLATFORM_ID = 'claude-code';
+export const PLATFORM_LABEL = 'Claude Code';
+
+export async function install({ tasDir, target }) {
+  const claudeDir = path.join(target, '.claude');
+  await fs.mkdir(claudeDir, { recursive: true });
+
+  await copyDir(path.join(tasDir, 'commands'), path.join(claudeDir, 'commands'));
+  console.log('  [ok] .claude/commands/');
+
+  await copyDir(path.join(tasDir, 'agents'), path.join(claudeDir, 'agents'));
+  console.log('  [ok] .claude/agents/');
+
+  const skillsSrc = path.join(tasDir, 'skills');
+  if (await exists(skillsSrc)) {
+    await copyDir(skillsSrc, path.join(claudeDir, 'skills'));
+    console.log('  [ok] .claude/skills/');
+  }
+
+  const hooksSrc = path.join(tasDir, '_platform', 'hooks');
+  if (await exists(hooksSrc)) {
+    await copyDir(hooksSrc, path.join(claudeDir, 'hooks'));
+    console.log('  [ok] .claude/hooks/');
+  }
+
+  const settingsSrc = path.join(tasDir, '_platform', 'claude-code', 'settings.json');
+  if (await exists(settingsSrc)) {
+    await fs.copyFile(settingsSrc, path.join(claudeDir, 'settings.json'));
+    console.log('  [ok] .claude/settings.json');
+  }
+}