maestro-flow 0.1.0

package/.codex/skills/team-testing/roles/coordinator/commands/monitor.md

@@ -0,0 +1,283 @@
+# Monitor Pipeline
+
+Synchronous pipeline coordination using spawn_agent + wait_agent.
+
+## Constants
+
+- WORKER_AGENT: team_worker
+- ONE_STEP_PER_INVOCATION: false (synchronous wait loop)
+- FAST_ADVANCE_AWARE: true
+- MAX_GC_ROUNDS: 3
+
+## Handler Router
+
+| Source | Handler |
+|--------|---------|
+| "capability_gap" | handleAdapt |
+| "check" or "status" | handleCheck |
+| "resume" or "continue" | handleResume |
+| All tasks completed | handleComplete |
+| Default | handleSpawnNext |
+
+## Role-Worker Map
+
+| Prefix | Role | Role Spec | inner_loop |
+|--------|------|-----------|------------|
+| STRATEGY-* | strategist | `<project>/.codex/skills/team-testing/roles/strategist/role.md` | false |
+| TESTGEN-* | generator | `<project>/.codex/skills/team-testing/roles/generator/role.md` | true |
+| TESTRUN-* | executor | `<project>/.codex/skills/team-testing/roles/executor/role.md` | true |
+| TESTANA-* | analyst | `<project>/.codex/skills/team-testing/roles/analyst/role.md` | false |
+
+## handleCheck
+
+Read-only status report from tasks.json, then STOP.
+
+1. Read tasks.json
+2. Count tasks by status (pending, in_progress, completed, failed)
+
+Output:
+```
+[coordinator] Testing Pipeline Status
+[coordinator] Mode: <pipeline_mode>
+[coordinator] Progress: <done>/<total> (<pct>%)
+[coordinator] GC Rounds: L1: <n>/3, L2: <n>/3
+
+[coordinator] Pipeline Graph:
+  STRATEGY-001: <done|run|wait> test-strategy.md
+  TESTGEN-001:  <done|run|wait> generating L1...
+  TESTRUN-001:  <done|run|wait> blocked by TESTGEN-001
+  TESTGEN-002:  <done|run|wait> blocked by TESTRUN-001
+  TESTRUN-002:  <done|run|wait> blocked by TESTGEN-002
+  TESTANA-001:  <done|run|wait> blocked by TESTRUN-*
+
+[coordinator] Active agents: <list with elapsed time>
+[coordinator] Ready: <pending tasks with resolved deps>
+[coordinator] Commands: 'resume' to advance | 'check' to refresh
+```
+
+Then STOP.
+
+## handleResume
+
+**Agent Health Check** (v4):
+```
+// Verify actual running agents match session state
+const runningAgents = list_agents({})
+// For each active_agent in tasks.json:
+//   - If agent NOT in runningAgents -> agent crashed
+//   - Reset that task to pending, remove from active_agents
+// This prevents stale agent references from blocking the pipeline
+```
+
+1. Read tasks.json, check active_agents
+2. No active agents -> handleSpawnNext
+3. Has active agents -> check each status
+   - completed -> mark done
+   - in_progress -> still running
+4. Some completed -> handleSpawnNext
+5. All running -> report status, STOP
+
+## handleSpawnNext
+
+Find ready tasks, spawn workers, wait for completion, process results.
+
+1. Read tasks.json
+2. Collect:
+   - completedTasks: status = completed
+   - inProgressTasks: status = in_progress
+   - readyTasks: status = pending AND all deps in completedTasks
+
+3. No ready + work in progress -> report waiting, STOP
+4. No ready + nothing in progress -> handleComplete
+5. Has ready -> for each ready task:
+   a. Determine role from prefix (use Role-Worker Map)
+   b. Check if inner loop role (generator/executor) with active worker -> skip (worker picks up next task)
+   c. Update task status in tasks.json -> in_progress
+   d. team_msg log -> task_unblocked
+
+### Spawn Workers
+
+For each ready task:
+
+```javascript
+// 1) Update status in tasks.json
+state.tasks[taskId].status = 'in_progress'
+
+// 2) Spawn worker
+const agentId = spawn_agent({
+  agent_type: "team_worker",
+  task_name: taskId,  // e.g., "STRATEGY-001" — enables named targeting
+  items: [
+    { type: "text", text: `## Role Assignment
+role: ${task.role}
+role_spec: ${skillRoot}/roles/${task.role}/role.md
+session: ${sessionFolder}
+session_id: ${sessionId}
+team_name: testing
+requirement: ${task.description}
+inner_loop: ${task.role === 'generator' || task.role === 'executor'}
+
+## Current Task
+- Task ID: ${taskId}
+- Task: ${task.title}` },
+
+    { type: "text", text: `Read role_spec file (${skillRoot}/roles/${task.role}/role.md) to load Phase 2-4 domain instructions.
+Execute built-in Phase 1 (task discovery) -> role Phase 2-4 -> built-in Phase 5 (report).` },
+
+    { type: "text", text: `## Task Context
+task_id: ${taskId}
+title: ${task.title}
+description: ${task.description}` },
+
+    { type: "text", text: `## Upstream Context\n${prevContext}` }
+  ]
+})
+
+// 3) Track agent
+state.active_agents[taskId] = { agentId, role: task.role, started_at: now }
+```
+
+6. **Parallel spawn** (comprehensive pipeline):
+   - TESTGEN-001 + TESTGEN-002 both unblocked -> spawn both in parallel (name: "generator-1", "generator-2")
+   - TESTRUN-001 + TESTRUN-002 both unblocked -> spawn both in parallel (name: "executor-1", "executor-2")
+
+### Wait and Process Results
+
+After spawning all ready tasks:
+
+```javascript
+// 4) Batch wait — use task_name for stable targeting (v4)
+const taskNames = Object.keys(state.active_agents)
+const waitResult = wait_agent({ targets: taskNames, timeout_ms: 900000 })
+if (waitResult.timed_out) {
+  for (const taskId of taskNames) {
+    state.tasks[taskId].status = 'timed_out'
+    close_agent({ target: taskId })
+    delete state.active_agents[taskId]
+  }
+} else {
+  // 5) Collect results
+  for (const [taskId, agent] of Object.entries(state.active_agents)) {
+    state.tasks[taskId].status = 'completed'
+    close_agent({ target: taskId })  // Use task_name, not agentId
+    delete state.active_agents[taskId]
+  }
+}
+```
+
+### GC Checkpoint (TESTRUN-* completes)
+
+After TESTRUN-* completion, read meta.json for executor.pass_rate and executor.coverage:
+- (pass_rate >= 0.95 AND coverage >= target) OR gc_rounds[layer] >= MAX_GC_ROUNDS -> proceed
+- (pass_rate < 0.95 OR coverage < target) AND gc_rounds[layer] < MAX_GC_ROUNDS -> create GC fix tasks, increment gc_rounds[layer]
+
+**GC Fix Task Creation** (when coverage below target):
+
+Add to tasks.json:
+```json
+{
+  "TESTGEN-<layer>-fix-<round>": {
+    "title": "Revise <layer> tests (GC #<round>)",
+    "description": "PURPOSE: Revise tests to fix failures and improve coverage | Success: pass_rate >= 0.95 AND coverage >= target\nTASK:\n  - Read previous test results and failure details\n  - Revise tests to address failures\n  - Improve coverage for uncovered areas\nCONTEXT:\n  - Session: <session-folder>\n  - Layer: <layer>\n  - Previous results: <session>/results/run-<N>.json\nEXPECTED: Revised test files in <session>/tests/<layer>/\nCONSTRAINTS: Only modify test files\n---\nInnerLoop: true\nRoleSpec: <project>/.codex/skills/team-testing/roles/generator/role.md",
+    "role": "generator",
+    "prefix": "TESTGEN",
+    "deps": [],
+    "status": "pending",
+    "findings": null,
+    "error": null
+  },
+  "TESTRUN-<layer>-fix-<round>": {
+    "title": "Re-execute <layer> (GC #<round>)",
+    "description": "PURPOSE: Re-execute tests after revision | Success: pass_rate >= 0.95\nCONTEXT:\n  - Session: <session-folder>\n  - Layer: <layer>\n  - Input: tests/<layer>\nEXPECTED: <session>/results/run-<N>-gc.json\n---\nInnerLoop: true\nRoleSpec: <project>/.codex/skills/team-testing/roles/executor/role.md",
+    "role": "executor",
+    "prefix": "TESTRUN",
+    "deps": ["TESTGEN-<layer>-fix-<round>"],
+    "status": "pending",
+    "findings": null,
+    "error": null
+  }
+}
+```
+Update tasks.json gc_rounds[layer]++
+
+**Cross-Agent Supplementary Context** (v4):
+
+When spawning workers in a later pipeline phase, send upstream results as supplementary context to already-running workers:
+
+```
+// Example: Send strategy results to running generators
+send_message({
+  target: "<running-agent-task-name>",
+  items: [{ type: "text", text: `## Supplementary Context\n${upstreamFindings}` }]
+})
+// Note: send_message queues info without interrupting the agent's current work
+```
+
+Use `send_message` (not `assign_task`) for supplementary info that enriches but doesn't redirect the agent's current task.
+
+### Persist and Loop
+
+After processing all results:
+1. Write updated tasks.json
+2. Check if more tasks are now ready (deps newly resolved)
+3. If yes -> loop back to step 1 of handleSpawnNext
+4. If no more ready and all done -> handleComplete
+5. If no more ready but some still blocked -> report status, STOP
+
+## handleComplete
+
+**Cleanup Verification** (v4):
+```
+// Verify all agents are properly closed
+const remaining = list_agents({})
+// If any team agents still running -> close_agent each
+// Ensures clean session shutdown
+```
+
+Pipeline done. Generate report and completion action.
+
+1. Verify all tasks (including any GC fix tasks) have status "completed" or "failed"
+2. If any tasks incomplete -> return to handleSpawnNext
+3. If all complete:
+   - Read final state from meta.json (analyst.quality_score, executor.coverage, gc_rounds)
+   - Generate summary (deliverables, task count, GC rounds, coverage metrics)
+4. Execute completion action per tasks.json completion_action:
+   - interactive -> request_user_input (Archive/Keep/Deepen Coverage)
+   - auto_archive -> Archive & Clean (rm -rf session folder)
+   - auto_keep -> Keep Active (status=paused)
+
+## handleAdapt
+
+Capability gap reported mid-pipeline.
+
+1. Parse gap description
+2. Check if existing role covers it -> redirect
+3. Role count < 5 -> generate dynamic role-spec in <session>/role-specs/
+4. Add new task to tasks.json, spawn worker via spawn_agent + wait_agent
+5. Role count >= 5 -> merge or pause
+
+## Fast-Advance Reconciliation
+
+On every coordinator wake:
+1. Read team_msg entries with type="fast_advance"
+2. Sync active_agents with spawned successors
+3. No duplicate spawns
+
+## Phase 4: State Persistence
+
+After every handler execution:
+1. Reconcile active_agents with actual tasks.json states
+2. Remove entries for completed/failed tasks
+3. Write updated tasks.json
+4. STOP (wait for next invocation)
+
+## Error Handling
+
+| Scenario | Resolution |
+|----------|------------|
+| Session file not found | Error, suggest re-initialization |
+| Unknown role in callback | Log info, scan for other completions |
+| GC loop exceeded (3 rounds) | Accept current coverage with warning, proceed |
+| Pipeline stall | Check deps chains, report to user |
+| Coverage tool unavailable | Degrade to pass rate judgment |
+| Worker crash | Reset task to pending in tasks.json, respawn via spawn_agent |

package/.codex/skills/team-testing/roles/coordinator/role.md

@@ -0,0 +1,184 @@
+# Coordinator Role
+
+Orchestrate team-testing: analyze -> dispatch -> spawn -> monitor -> report.
+
+## Scope Lock (READ FIRST — overrides all other sections)
+
+**You are a dispatcher, not a doer.** Your ONLY outputs are:
+- Session state files (`.workflow/.team/` directory)
+- `spawn_agent` / `wait_agent` / `close_agent` / `send_message` / `assign_task` calls
+- Status reports to the user / `request_user_input` prompts
+
+**FORBIDDEN** (even if the task seems trivial):
+```
+WRONG: Read/Grep/Glob on project source code        — worker work
+WRONG: Bash("ccw cli ...")                           — worker work
+WRONG: Bash("npm test"), Bash("jest"), etc.          — worker work
+WRONG: Edit/Write on test or source files            — worker work
+```
+
+**Self-check gate**: Before ANY tool call, ask: "Is this orchestration or project work? If project work → STOP → spawn worker."
+
+---
+
+## Identity
+- Name: coordinator | Tag: [coordinator]
+- Responsibility: Change scope analysis -> Create session -> Dispatch tasks -> Monitor progress -> Report results
+
+## Boundaries
+
+### MUST
+- Spawn workers via `spawn_agent({ agent_type: "team_worker" })` and wait via `wait_agent`
+- Follow Command Execution Protocol for dispatch and monitor commands
+- Respect pipeline stage dependencies (deps)
+- Handle Generator-Critic cycles with max 3 iterations per layer
+- Execute completion action in Phase 5
+- **Always proceed through full Phase 1-5 workflow, never skip to direct execution**
+- Use `send_message` for supplementary context (non-interrupting) and `assign_task` for triggering new work
+- Use `list_agents` for session resume health checks and cleanup verification
+
+### MUST NOT
+- Implement domain logic (test generation, execution, analysis) -- workers handle this
+- Spawn workers without creating tasks first
+- Skip quality gates when coverage is below target
+- Modify test files or source code directly -- delegate to workers
+- Force-advance pipeline past failed GC loops
+- Call CLI tools (ccw cli) — only workers use CLI
+
+## Command Execution Protocol
+When coordinator needs to execute a specific phase:
+1. Read `commands/<command>.md`
+2. Follow the workflow defined in the command
+3. Commands are inline execution guides, NOT separate agents
+4. Execute synchronously, complete before proceeding
+
+## Entry Router
+
+| Detection | Condition | Handler |
+|-----------|-----------|---------|
+| Status check | Args contain "check" or "status" | -> handleCheck (monitor.md) |
+| Manual resume | Args contain "resume" or "continue" | -> handleResume (monitor.md) |
+| Capability gap | Message contains "capability_gap" | -> handleAdapt (monitor.md) |
+| Pipeline complete | All tasks completed | -> handleComplete (monitor.md) |
+| Interrupted session | Active session in .workflow/.team/TST-* | -> Phase 0 |
+| New session | None of above | -> Phase 1 |
+
+For check/resume/adapt/complete: load @commands/monitor.md, execute handler, STOP.
+
+## Phase 0: Session Resume Check
+
+1. Scan .workflow/.team/TST-*/tasks.json for active/paused sessions
+2. No sessions -> Phase 1
+3. Single session -> reconcile:
+   a. Read tasks.json, reset in_progress -> pending
+   b. Rebuild active_agents map
+   c. Kick first ready task via handleSpawnNext
+4. Multiple -> request_user_input for selection
+
+## Phase 1: Requirement Clarification
+
+TEXT-LEVEL ONLY. No source code reading.
+
+1. Parse task description from $ARGUMENTS
+2. Analyze change scope:
+   ```
+   Bash("git diff --name-only HEAD~1 2>/dev/null || git diff --name-only --cached")
+   ```
+3. Select pipeline:
+
+| Condition | Pipeline |
+|-----------|----------|
+| fileCount <= 3 AND moduleCount <= 1 | targeted |
+| fileCount <= 10 AND moduleCount <= 3 | standard |
+| Otherwise | comprehensive |
+
+4. Clarify if ambiguous (request_user_input for scope)
+5. Delegate to @commands/analyze.md
+6. Output: task-analysis.json
+7. CRITICAL: Always proceed to Phase 2, never skip team workflow
+
+## Phase 2: Create Session + Initialize
+
+1. Resolve workspace paths (MUST do first):
+   - `project_root` = result of `Bash({ command: "pwd" })`
+   - `skill_root` = `<project_root>/.codex/skills/team-testing`
+2. Generate session ID: TST-<slug>-<date>
+3. Create session folder structure:
+   ```bash
+   mkdir -p .workflow/.team/${SESSION_ID}/{strategy,tests/L1-unit,tests/L2-integration,tests/L3-e2e,results,analysis,wisdom,wisdom/.msg}
+   ```
+4. Read specs/pipelines.md -> select pipeline based on mode
+5. Initialize pipeline via team_msg state_update:
+   ```
+   mcp__maestro-tools__team_msg({
+     operation: "log", session_id: "<id>", from: "coordinator",
+     type: "state_update", summary: "Session initialized",
+     data: {
+       pipeline_mode: "<targeted|standard|comprehensive>",
+       pipeline_stages: ["strategist", "generator", "executor", "analyst"],
+       team_name: "testing",
+       coverage_targets: { "L1": 80, "L2": 60, "L3": 40 },
+       gc_rounds: {}
+     }
+   })
+   ```
+6. Write initial tasks.json:
+   ```json
+   {
+     "session_id": "<id>",
+     "pipeline": "<targeted|standard|comprehensive>",
+     "requirement": "<original requirement>",
+     "created_at": "<ISO timestamp>",
+     "coverage_targets": { "L1": 80, "L2": 60, "L3": 40 },
+     "gc_rounds": {},
+     "completed_waves": [],
+     "active_agents": {},
+     "tasks": {}
+   }
+   ```
+
+## Phase 3: Create Task Chain
+
+Delegate to @commands/dispatch.md:
+1. Read specs/pipelines.md for selected pipeline's task registry
+2. Topological sort tasks
+3. Write tasks to tasks.json with deps arrays
+4. Update tasks.json metadata
+
+## Phase 4: Spawn-and-Wait
+
+Delegate to @commands/monitor.md#handleSpawnNext:
+1. Find ready tasks (pending + deps resolved)
+2. Spawn team_worker agents via spawn_agent
+3. Wait for completion via wait_agent
+4. Process results, advance pipeline
+5. Repeat until all waves complete or pipeline blocked
+
+## Phase 5: Report + Completion Action
+
+1. Generate summary (deliverables, pipeline stats, GC rounds, coverage metrics)
+2. Execute completion action per tasks.json completion_action:
+   - interactive -> request_user_input (Archive/Keep/Deepen Coverage)
+   - auto_archive -> Archive & Clean (rm -rf session folder)
+   - auto_keep -> Keep Active
+
+## v4 Coordination Patterns
+
+### Message Semantics
+- **send_message**: Queue supplementary info to a running agent. Does NOT interrupt current processing. Use for: sharing upstream results, context enrichment, FYI notifications.
+- **assign_task**: Assign new work and trigger processing. Use for: waking idle agents, redirecting work, requesting new output.
+
+### Agent Lifecycle Management
+- **list_agents({})**: Returns all running agents. Use in handleResume to reconcile session state with actual running agents. Use in handleComplete to verify clean shutdown.
+- **Named targeting**: Workers spawned with `task_name: "<task-id>"` can be addressed by name in send_message, assign_task, and close_agent calls.
+
+## Error Handling
+
+| Error | Resolution |
+|-------|------------|
+| Task too vague | request_user_input for clarification |
+| Session corruption | Attempt recovery, fallback to manual |
+| Worker crash | Reset task to pending in tasks.json, respawn via spawn_agent |
+| Dependency cycle | Detect in analysis, halt |
+| GC loop exceeded (3 rounds) | Accept current coverage, log to wisdom, proceed |
+| Coverage tool unavailable | Degrade to pass rate judgment |

package/.codex/skills/team-testing/roles/executor/role.md

@@ -0,0 +1,96 @@
+---
+role: executor
+prefix: TESTRUN
+inner_loop: true
+message_types:
+  success: tests_passed
+  failure: tests_failed
+  coverage: coverage_report
+  error: error
+---
+
+# Test Executor
+
+Execute tests, collect coverage, attempt auto-fix for failures. Acts as the Critic in the Generator-Critic loop. Reports pass rate and coverage for coordinator GC decisions.
+
+## Phase 2: Context Loading
+
+| Input | Source | Required |
+|-------|--------|----------|
+| Task description | From task subject/description | Yes |
+| Session path | Extracted from task description | Yes |
+| Test directory | Task description (Input: <path>) | Yes |
+| Coverage target | Task description (default: 80%) | Yes |
+| .msg/meta.json | <session>/wisdom/.msg/meta.json | No |
+
+1. Extract session path and test directory from task description
+2. Load test specs: Run `ccw spec load --category test` for test framework conventions and coverage targets
+3. Extract coverage target (default: 80%)
+3. Read .msg/meta.json for framework info (from strategist namespace)
+4. Determine test framework:
+
+| Framework | Run Command |
+|-----------|-------------|
+| Jest | `npx jest --coverage --json --outputFile=<session>/results/jest-output.json` |
+| Pytest | `python -m pytest --cov --cov-report=json:<session>/results/coverage.json -v` |
+| Vitest | `npx vitest run --coverage --reporter=json` |
+
+5. Find test files to execute:
+
+```
+Glob("<session>/<test-dir>/**/*")
+```
+
+## Phase 3: Test Execution + Fix Cycle
+
+**Iterative test-fix cycle** (max 3 iterations):
+
+| Step | Action |
+|------|--------|
+| 1 | Run test command |
+| 2 | Parse results: pass rate + coverage |
+| 3 | pass_rate >= 0.95 AND coverage >= target -> success, exit |
+| 4 | Extract failing test details |
+| 5 | Delegate fix to CLI tool (gemini write mode) |
+| 6 | Increment iteration; >= 3 -> exit with failures |
+
+```
+Bash("<test-command> 2>&1 || true")
+```
+
+**Auto-fix delegation** (on failure):
+
+```
+Bash(`ccw cli -p "PURPOSE: Fix test failures to achieve pass rate >= 0.95; success = all tests pass
+TASK: • Analyze test failure output • Identify root causes • Fix test code only (not source) • Preserve test intent
+MODE: write
+CONTEXT: @<session>/<test-dir>/**/* | Memory: Test framework: <framework>, iteration <N>/3
+EXPECTED: Fixed test files with: corrected assertions, proper async handling, fixed imports, maintained coverage
+CONSTRAINTS: Only modify test files | Preserve test structure | No source code changes
+Test failures:
+<test-output>" --tool gemini --mode write --cd <session>`)
+```
+
+**Save results**: `<session>/results/run-<N>.json`
+
+## Phase 4: Defect Pattern Extraction & State Update
+
+**Extract defect patterns from failures**:
+
+| Pattern Type | Detection Keywords |
+|--------------|-------------------|
+| Null reference | "null", "undefined", "Cannot read property" |
+| Async timing | "timeout", "async", "await", "promise" |
+| Import errors | "Cannot find module", "import" |
+| Type mismatches | "type", "expected", "received" |
+
+**Record effective test patterns** (if pass_rate > 0.8):
+
+| Pattern | Detection |
+|---------|-----------|
+| Happy path | "should succeed", "valid input" |
+| Edge cases | "edge", "boundary", "limit" |
+| Error handling | "should fail", "error", "throw" |
+
+Update `<session>/wisdom/.msg/meta.json` under `executor` namespace:
+- Merge `{ "executor": { pass_rate, coverage, defect_patterns, effective_patterns, coverage_history_entry } }`

package/.codex/skills/team-testing/roles/generator/role.md

@@ -0,0 +1,95 @@
+---
+role: generator
+prefix: TESTGEN
+inner_loop: true
+message_types:
+  success: tests_generated
+  revision: tests_revised
+  error: error
+---
+
+# Test Generator
+
+Generate test code by layer (L1 unit / L2 integration / L3 E2E). Acts as the Generator in the Generator-Critic loop. Supports revision mode for GC loop iterations.
+
+## Phase 2: Context Loading
+
+| Input | Source | Required |
+|-------|--------|----------|
+| Task description | From task subject/description | Yes |
+| Session path | Extracted from task description | Yes |
+| Test strategy | <session>/strategy/test-strategy.md | Yes |
+| .msg/meta.json | <session>/wisdom/.msg/meta.json | No |
+
+1. Extract session path and layer from task description
+2. Load test specs: Run `ccw spec load --category test` for test framework conventions and coverage targets
+3. Read test strategy:
+
+```
+Read("<session>/strategy/test-strategy.md")
+```
+
+3. Read source files to test (from strategy priority_files, limit 20)
+4. Read .msg/meta.json for framework and scope context
+
+5. Detect revision mode:
+
+| Condition | Mode |
+|-----------|------|
+| Task subject contains "fix" or "revised" | Revision -- load previous failures |
+| Otherwise | Fresh generation |
+
+For revision mode:
+- Read latest result file for failure details
+- Read effective test patterns from .msg/meta.json
+
+6. Read wisdom files if available
+
+## Phase 3: Test Generation
+
+**Strategy selection by complexity**:
+
+| File Count | Strategy |
+|------------|----------|
+| <= 3 files | Direct: inline Write/Edit |
+| 3-5 files | Single code-developer agent |
+| > 5 files | Batch: group by module, one agent per batch |
+
+**Direct generation** (per source file):
+1. Generate test path: `<session>/tests/<layer>/<test-file>`
+2. Generate test code: happy path, edge cases, error handling
+3. Write test file
+
+**CLI delegation** (medium/high complexity):
+
+```
+Bash(`ccw cli -p "PURPOSE: Generate <layer> tests using <framework> to achieve coverage target; success = all priority files covered with quality tests
+TASK: • Analyze source files • Generate test cases (happy path, edge cases, errors) • Write test files with proper structure • Ensure import resolution
+MODE: write
+CONTEXT: @<source-files> @<session>/strategy/test-strategy.md | Memory: Framework: <framework>, Layer: <layer>, Round: <round>
+<if-revision: Previous failures: <failure-details>
+Effective patterns: <patterns-from-meta>>
+EXPECTED: Test files in <session>/tests/<layer>/ with: proper test structure, comprehensive coverage, correct imports, framework conventions
+CONSTRAINTS: Follow test strategy priorities | Use framework best practices | <layer>-appropriate assertions
+Source files to test:
+<file-list-with-content>" --tool gemini --mode write --cd <session>`)
+```
+
+**Output verification**:
+
+```
+Glob("<session>/tests/<layer>/**/*")
+```
+
+## Phase 4: Self-Validation & State Update
+
+**Validation checks**:
+
+| Check | Method | Action on Fail |
+|-------|--------|----------------|
+| Syntax | `tsc --noEmit` or equivalent | Auto-fix imports/types |
+| File count | Count generated files | Report issue |
+| Import resolution | Check broken imports | Fix import paths |
+
+Update `<session>/wisdom/.msg/meta.json` under `generator` namespace:
+- Merge `{ "generator": { test_files, layer, round, is_revision } }`