aida

AIDA

Generate a complete project using multi-agent orchestration with TDD and quality gates.

Usage

/aida "Create a Twitter clone with Go backend and React frontend"

---

MANDATORY EXECUTION PROTOCOL

You MUST follow this protocol exactly. Do NOT deviate.

This command combines /aida:init + /aida:start + quality verification.

---

What This Command Does

1. Auto-Init: Creates output directories and validates environment
2. Session Start: Initializes session state with full tracking
3. Spec Generation: Launches Leader-Spec for phases 1-4 via Task tool
4. Implementation: Launches Leader-Impl for TDD implementation via Task tool
5. Quality Gates: Verifies all 18 quality gates pass (enforced by Stop Hook)
6. Completion: Reports final project location with verification

---

STRICT MODE (DEFAULT)

AIDA operates in strict mode by default:

Enforcement via Stop Hook

When AIDA attempts to complete, the Stop Hook intercepts and:

1. Runs ./scripts/quality-gates.sh automatically
2. Blocks exit if any gate fails (exit code 2)
3. Forces iteration until ALL requirements are met
4. Allows exit only when ALL gates pass (exit code 0)

Quality Requirements (MANDATORY)

Requirement	Minimum	Enforced By
Backend Tests	80+	Stop Hook
Frontend Tests	100+	Stop Hook
E2E Tests	20+ (actual execution)	Stop Hook + Gate 19
Backend Coverage	75%+	Stop Hook
Frontend Coverage	70%+	Stop Hook
Docker	Build/Run/Health/E2E	Quality Gates
TDD Evidence	10+ files	Gate 20

Completion Condition

"DONE" can ONLY be output when:

• All 20 quality gates PASS (including Gate 19: E2E, Gate 20: TDD Evidence)
• Stop Hook returns exit code 0
• quality_gates_passed: true in session.json
• E2E tests actually executed against running Docker containers
• TDD evidence recorded with RED-GREEN-REFACTOR cycle

Gate 19: E2E Test Execution

Gate 19 は Docker が起動中に Playwright E2E テストを実際に実行:

Gate 6: Docker Run → Gate 7: Health Check → Gate 19: E2E Execution → Cleanup

E2Eテストファイルの存在だけでは不十分。実際に実行してPASSする必要がある。

Gate 20: TDD Evidence Verification

Gate 20 はTDDサイクル（RED-GREEN-REFACTOR）の証拠を検証:

# TDDサイクルを記録
./scripts/tdd-logger.sh start <feature>
./scripts/tdd-logger.sh red <test-file>
./scripts/tdd-logger.sh green <test-file>
./scripts/tdd-logger.sh refactor "<changes>"
./scripts/tdd-logger.sh complete

証拠は .aida/tdd-evidence/ に保存。10+ ファイル必須。

No exceptions. No shortcuts. No manual overrides.

---

Step 1: Auto-Initialize

Create all required directories:

mkdir -p .aida/state .aida/checkpoints .aida/artifacts/requirements .aida/artifacts/designs .aida/tasks .aida/results .aida/specs .aida/errors .aida/tdd-evidence

Derive project name from user request:

• Convert to kebab-case
• Maximum 20 characters
• Examples:
  • "Create a Twitter clone" → twitter-clone
  • "Build todo app with auth" → todo-app
  • "Simple notes application" → notes-app

---

Step 2: Create Session

Create .aida/state/session.json:

{
  "session_id": "<UUID>",
  "started_at": "<ISO8601>",
  "mode": "aida",
  "current_phase": "SPEC_PHASE",
  "phase": 1,
  "phase_name": "extraction",
  "user_request": "$ARGUMENTS",
  "project_name": "<derived>",
  "phase_history": [
    {"phase": "INITIALIZING", "entered_at": "<ISO8601>", "exited_at": "<ISO8601>"}
  ],
  "leaders": {
    "spec": "pending",
    "impl": "pending"
  },
  "active_agents": [],
  "completed_tasks": [],
  "pending_tasks": ["spec-requirements", "spec-design", "spec-tasks", "impl-backend", "impl-frontend", "impl-docker", "quality-gates"]
}

Create .aida/kanban.md:

# Project Kanban - {{PROJECT_NAME}}

## Current Status: SPEC_PHASE (Phase 1)

## Spec Phase
- [ ] Phase 1: Extraction & Architecture
- [ ] Phase 2: Structure & Schema
- [ ] Phase 3: Alignment
- [ ] Phase 4: Verification

## Impl Phase
- [ ] Backend Implementation (TDD)
- [ ] Frontend Implementation (TDD)
- [ ] Docker Setup

## Quality Gates (ALL 19 MUST PASS)
- [ ] Gate 1: Backend Build
- [ ] Gate 2: Backend Tests
- [ ] Gate 3: Frontend Build
- [ ] Gate 4: Frontend Tests
- [ ] Gate 5: Docker Build
- [ ] Gate 6: Docker Run
- [ ] Gate 7: Health Check
- [ ] Gate 19: E2E Test Execution (Playwright)

---

Step 3: Launch Leader-Spec (Phases 1-4)

<MANDATORY_ACTION id="launch-leader-spec">

YOU MUST INVOKE THE TASK TOOL NOW.

Do NOT just describe the Task tool call - actually execute it.

Use these exact parameters:

Parameter	Value
description	"Leader-Spec: Specification Phases 1-4"
subagent_type	"general-purpose"
model	"sonnet"
run_in_background	false
prompt	See below

Task Prompt:

You are AIDA Leader-Spec agent.

## CRITICAL INSTRUCTION
Read and follow the full instructions in: agents/leader-spec.md

## Current Session
- Session ID: {{SESSION_ID}}
- Project: {{PROJECT_NAME}}
- User Request: {{USER_REQUEST}}
- Working Directory: {{CWD}}

## Your Mission

Execute Phases 1-4 of the AIDA pipeline:

### Phase 1: Extraction & Architecture
1. Analyze user requirements thoroughly
2. Extract core features and constraints
3. Design high-level architecture
4. Write .aida/artifacts/requirements/extraction.md

### Phase 2: Structure
1. Define directory structure
2. Create data schemas
3. Define API contracts
4. Write .aida/artifacts/designs/structure.md

### Phase 3: Alignment
1. Verify requirements consistency
2. Check for conflicts or gaps
3. Write .aida/artifacts/alignment.md

### Phase 4: Verification & Output
1. Review all specs for completeness
2. Write final specifications:
   - .aida/specs/{{PROJECT}}-requirements.md (comprehensive, min 500 bytes)
   - .aida/specs/{{PROJECT}}-design.md (technical design, min 500 bytes)
   - .aida/specs/{{PROJECT}}-tasks.md (implementation tasks)

## Player Delegation
For parallel tasks, spawn player subagents using Task tool:
- subagent_type: "general-purpose"
- model: "haiku"
- Read agents/player.md for player protocol

## Completion Checklist
Before completing, verify:
- [ ] .aida/specs/{{PROJECT}}-requirements.md exists (min 500 bytes)
- [ ] .aida/specs/{{PROJECT}}-design.md exists (min 500 bytes)
- [ ] .aida/specs/{{PROJECT}}-tasks.md exists

## Completion Report
Write to .aida/results/spec-complete.json:
{
  "task_id": "spec-{{PROJECT}}",
  "status": "completed",
  "completed_at": "ISO8601",
  "outputs": {
    "requirements": ".aida/specs/{{PROJECT}}-requirements.md",
    "design": ".aida/specs/{{PROJECT}}-design.md",
    "tasks": ".aida/specs/{{PROJECT}}-tasks.md"
  },
  "summary": "Specification phases 1-4 complete"
}

Update .aida/state/session.json with:
- current_phase: "IMPL_PHASE"
- phase: 5
- leaders.spec: "completed"

</MANDATORY_ACTION>

STOP: Do NOT proceed to Step 4 until Task tool has been invoked and Leader-Spec completes.

---

Step 4: Validate Specs

After Leader-Spec completes, validate:

./scripts/validate-outputs.sh {{PROJECT}} spec

If validation fails:

1. Report which files are missing
2. Re-spawn Leader-Spec to complete
3. Do NOT proceed until specs are valid

If validation passes:

• Continue to Step 5

---

Step 5: Launch Leader-Impl (Phase 5)

<MANDATORY_ACTION id="launch-leader-impl">

YOU MUST INVOKE THE TASK TOOL NOW.

Do NOT just describe the Task tool call - actually execute it.

Use these exact parameters:

Parameter	Value
description	"Leader-Impl: TDD Implementation Phase"
subagent_type	"general-purpose"
model	"sonnet"
run_in_background	false
prompt	See below

Task Prompt:

You are AIDA Leader-Impl agent.

## CRITICAL INSTRUCTION
Read and follow the full instructions in: agents/leader-impl.md

## Current Session
- Session ID: {{SESSION_ID}}
- Project: {{PROJECT_NAME}}
- Working Directory: {{CWD}}

## Specifications (MUST READ)
- .aida/specs/{{PROJECT}}-requirements.md
- .aida/specs/{{PROJECT}}-design.md
- .aida/specs/{{PROJECT}}-tasks.md

## TDD Protocol (MANDATORY)
Every implementation MUST follow:
1. RED: Write failing test FIRST
2. GREEN: Minimal code to pass test
3. REFACTOR: Clean up while tests pass

NO code without tests. NO tests without running them.

## Player Delegation (MANDATORY - ALL THREE PLAYERS)

### Backend Player
- subagent_type: "general-purpose"
- model: "haiku"
- Must produce: backend/ (in project directory)
- Must have: minimum 5 test files (*_test.go)
- All tests MUST pass

### Frontend Player (MANDATORY - SEPARATE)
- subagent_type: "general-purpose"
- model: "haiku"
- Must initialize with: npm create vite@latest frontend -- --template react-ts
- Must produce: frontend/ (in project directory)
- Must have: minimum 3 test files (*.test.tsx)
- All tests MUST pass

### Docker Player
- subagent_type: "general-purpose"
- model: "haiku"
- Must produce: docker-compose.yml, Dockerfiles
- Use Podman-compatible image paths: docker.io/library/...

## Quality Gates (ALL MUST PASS)
After all players complete, run:
./scripts/quality-gates.sh {{PROJECT}}

Gates:
1. Backend Build: go build ./...
2. Backend Tests: go test ./...
3. Frontend Build: npm run build
4. Frontend Tests: npm test -- --run
5. Docker Build: docker compose build
6. Docker Run: docker compose up -d
7. Health Check: curl localhost:8080/health

## Completion Checklist
Before completing:
- [ ] Backend directory has working Go code
- [ ] Backend has minimum 5 test files
- [ ] Frontend directory has working React code (NOT EMPTY)
- [ ] Frontend has minimum 3 test files
- [ ] Docker compose works
- [ ] ALL quality gates pass

## Completion Report
Write to .aida/results/impl-complete.json:
{
  "task_id": "impl-{{PROJECT}}",
  "status": "completed",
  "completed_at": "ISO8601",
  "project_path": "./",
  "quality_gates": {
    "backend_build": true,
    "backend_tests": true,
    "frontend_build": true,
    "frontend_tests": true,
    "docker_build": true,
    "docker_run": true,
    "health_check": true,
    "all_passed": true
  },
  "verification": {
    "backend": {"test_count": N, "test_output": "..."},
    "frontend": {"test_count": N, "test_output": "..."}
  },
  "summary": "Implementation complete, all quality gates passed"
}

Update .aida/state/session.json:
- current_phase: "COMPLETED"
- leaders.impl: "completed"

</MANDATORY_ACTION>

STOP: Wait for Task tool completion before proceeding.

---

Step 6: Run Quality Gates

After Leader-Impl completes, run full verification:

./scripts/quality-gates.sh {{PROJECT}}

All 7 gates MUST pass:

1. Backend Build
2. Backend Tests
3. Frontend Build
4. Frontend Tests
5. Docker Build
6. Docker Run
7. Health Check

If any gate fails:

1. Identify the failure
2. Fix or re-spawn appropriate player
3. Re-run gates until all pass

---

Step 7: Report Completion

After all quality gates pass:

Update .aida/kanban.md:

# Project Kanban - {{PROJECT_NAME}}

## Status: COMPLETED

## Spec Phase - COMPLETE
- [x] Phase 1: Extraction & Architecture
- [x] Phase 2: Structure & Schema
- [x] Phase 3: Alignment
- [x] Phase 4: Verification

## Impl Phase - COMPLETE
- [x] Backend Implementation (TDD)
- [x] Frontend Implementation (TDD)
- [x] Docker Setup

## Quality Gates - ALL PASSED
- [x] Gate 1: Backend Build
- [x] Gate 2: Backend Tests
- [x] Gate 3: Frontend Build
- [x] Gate 4: Frontend Tests
- [x] Gate 5: Docker Build
- [x] Gate 6: Docker Run
- [x] Gate 7: Health Check
- [x] Gate 19: E2E Test Execution (Playwright)

Final Output:

AIDA Complete

Session: {{SESSION_ID}}
Project: {{PROJECT_NAME}}
Duration: {{DURATION}}

Generated Artifacts:
- Specs: .aida/specs/{{PROJECT}}-*.md
- Project: ./

Quality Gates: 7/7 PASSED
- Backend Build: PASS
- Backend Tests: PASS
- Frontend Build: PASS
- Frontend Tests: PASS
- Docker Build: PASS
- Docker Run: PASS
- Health Check: PASS

TDD Verification:
- Backend: {{N}} test files, all passing
- Frontend: {{N}} test files, all passing

To run the project:
  docker compose up -d
  open http://localhost:5173

To verify quality gates again:
  ./scripts/quality-gates.sh

---

Multi-Agent Architecture

/aida "Create X"
    |
    +-- Step 1: Auto-Initialize (directories)
    |
    +-- Step 2: Create Session (session.json, kanban.md)
    |
    +-- Step 3: Task tool (sonnet) --> [Leader-Spec]
    |                                      |
    |                                      +-- Task tool (haiku) --> [Player]
    |                                      +-- Task tool (haiku) --> [Player]
    |                                      |
    |                                      +--> .aida/specs/
    |
    +-- Step 4: Validate Specs (validate-outputs.sh)
    |
    +-- Step 5: Task tool (sonnet) --> [Leader-Impl]
    |                                      |
    |                                      +-- Task tool (haiku) --> [Backend Player]
    |                                      +-- Task tool (haiku) --> [Frontend Player]
    |                                      +-- Task tool (haiku) --> [Docker Player]
    |                                      |
    |                                      +--> ./
    |
    +-- Step 6: Quality Gates (quality-gates.sh)
    |       |
    |       +-- 7 mandatory gates
    |
    +-- Step 7: Report Completion

---

CRITICAL REQUIREMENTS

1. Task tool MUST be invoked - Leaders run as subagents via Task tool
2. Wait for completion - run_in_background: false ensures sequential execution
3. Verify outputs exist - Check spec files were actually created
4. All quality gates MUST pass - No success without 7/7
5. TDD mandatory - No code without tests
6. Frontend SEPARATE - Must spawn dedicated Frontend Player
7. Model selection - Leaders: sonnet, Players: haiku

---

Status Check

To check progress during or after execution:

/aida:status

---

Validation Commands

# Validate spec outputs
./scripts/validate-outputs.sh {{PROJECT}} spec

# Validate impl outputs
./scripts/validate-outputs.sh {{PROJECT}} impl

# Verify TDD compliance
./scripts/verify-tdd.sh {{PROJECT}} all

# Run all quality gates
./scripts/quality-gates.sh {{PROJECT}}

---

Related Commands

Command	Description
/aida:init	Initialize directories only
/aida:start	Start spec phase only
/aida:work	Continue current phase
/aida:status	Check current status
/aida:pipeline	Full automation (same as /aida)
/aida:resume	Continue from last session state
/aida:fix <project>	Fix existing project to meet all quality gates