plan-parallel（平行计划）

[IMPORTANT] Use TaskCreate to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI MUST ATTENTION ask user whether to skip.

Understand Code First — HARD-GATE: Do NOT write, plan, or fix until you READ existing code. 1. Search 3+ similar patterns (grep/glob) — cite file:line evidence 2. Read existing files in target area — understand structure, base classes, conventions 3. Run python.claude/scripts/code_graph trace <file> --direction both --json when .code-graph/graph.db exists 4. Map dependencies via connections or callers_of — know what depends on your target 5. Write investigation to .ai/workspace/analysis/ for non-trivial tasks (3+ files) 6. Re-read analysis file before implementing — never work from memory alone 7. NEVER invent new patterns when existing ones work — match exactly or document deviation BLOCKED until: - [] Read target files - [] Grep 3+ patterns - [] Graph trace (if graph.db exists) - [] Assumptions verified with evidence

Estimation — Modified Fibonacci: 1(trivial) → 2(small) → 3(medium) → 5(large) → 8(very large) → 13(epic, SHOULD split) → 21(MUST ATTENTION split). Output story_points and complexity in plan frontmatter. Complexity auto-derived: 1-2=Low, 3-5=Medium, 8=High, 13+=Critical.

Plan Quality — Every plan phase MUST ATTENTION include test specifications. 1. Add ## Test Specifications section with TC-{FEAT}-{NNN} IDs to every phase file 2. Map every functional requirement to ≥1 TC (or explicit TBD with rationale) 3. TC IDs follow TC-{FEATURE}-{NNN} format — reference by ID, never embed full content 4. Before any new workflow step: call TaskList and re-read the phase file 5. On context compaction: call TaskList FIRST — never create duplicate tasks 6. Verify TC satisfaction per phase before marking complete (evidence must be file:line, not TBD) Mode: TDD-first → reference existing TCs with Evidence: TBD. Implement-first → use TBD → /tdd-spec fills after.

Iterative Phase Quality — Score complexity BEFORE planning. Complexity signals: >5 files +2, cross-service +3, new pattern +2, DB migration +2 Score >=6 → MUST ATTENTION decompose into phases. Each phase: - ≤5 files modified - ≤3h effort - Follows cycle: plan → implement → review → fix → verify - Do NOT start Phase N+1 until Phase N passes VERIFY Phase success = all TCs pass + code-reviewer agent approves + no CRITICAL findings.

• docs/test-specs/ — Test specifications by module (read existing TCs to include test strategy in plan)

Quick Summary

Goal: Create a detailed implementation plan with phases optimized for parallel execution by multiple agents.

Workflow:

1. Research — Spawn up to 2 researcher agents in parallel for different aspects
2. Analyze Codebase — Read summary docs; scout if summary is stale (>3 days)
3. Plan Creation — Planner subagent creates parallel-optimized phases with exclusive file ownership
4. User Review — Present plan for confirmation; optionally validate via interview

Key Rules:

• Each phase MUST ATTENTION have exclusive file ownership -- no file modified by multiple phases
• Phases must be independently executable with clear dependency matrix
• Planning-only skill -- never implement code, always get user approval first

Be skeptical. Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence percentages (Idea should be more than 80%).

Activate planning skill.

PLANNING-ONLY — Collaboration Required

DO NOT use the EnterPlanMode tool — you are ALREADY in a planning workflow. DO NOT implement or execute any code changes. COLLABORATE with the user: ask decision questions, present options with recommendations. After plan creation, ALWAYS run /plan-review to validate the plan. ASK user to confirm the plan before any next step.

Your mission

Workflow

1. Create a directory using naming pattern from ## Naming section in injected context. Make sure you pass the directory path to every subagent during the process.
2. Follow strictly to the "Plan Creation & Organization" rules of planning skill.
3. Use multiple researcher agents (max 2 agents) in parallel to research for this task: Each agent research for a different aspect of the task and are allowed to perform max 5 tool calls.
4. Analyze the codebase by reading backend-patterns-reference.md, frontend-patterns-reference.md, and project-structure-reference.md file. ONLY PERFORM THIS FOLLOWING STEP IF reference docs are placeholders or older than 3 days: Use /scout <instructions> slash command to search the codebase for files needed to complete the task.
5. Main agent gathers all research and scout report filepaths, and pass them to planner subagent with the prompt to create a parallel-optimized implementation plan.
6. Main agent receives the implementation plan from planner subagent, and ask user to review the plan

Post-Plan Validation (Optional)

After plan creation, offer validation interview to confirm decisions before implementation.

Check ## Plan Context -> Validation: mode=X, questions=MIN-MAX:

If mode is prompt: Use AskUserQuestion tool with options above. If user chooses validation or mode is auto: Execute /plan-validate {plan-path} SlashCommand.

Special Requirements for Parallel Execution

CRITICAL: The planner subagent must create phases that:

1. Can be executed independently - Each phase should be self-contained with no runtime dependencies on other phases
2. Have clear boundaries - No file overlap between phases (each file should only be modified in ONE phase)
3. Separate concerns logically - Group by architectural layer, feature domain, or technology stack
4. Minimize coupling - Phases should communicate through well-defined interfaces only
5. Include dependency matrix - Clearly document which phases must run sequentially vs in parallel

Parallelization Strategy:

• Group frontend/backend/database work into separate phases when possible
• Separate infrastructure setup from application logic
• Isolate different feature domains (e.g., auth vs profile vs payments)
• Split by file type/directory (e.g., components vs services vs models)
• Create independent test phases per module

Phase Organization Example:

Phase 01: Database Schema (can run independently)
Phase 02: Backend API Layer (can run independently)
Phase 03: Frontend Components (can run independently)
Phase 04: Integration Tests (depends on 01, 02, 03)

Output Requirements

Plan Directory Structure (use Plan dir: from ## Naming section)

{plan-dir}/
├── research/
│   ├── researcher-XX-report.md
│   └── ...
├── reports/
│   ├── XX-report.md
│   └── ...
├── scout/
│   ├── scout-XX-report.md
│   └── ...
├── plan.md
├── phase-XX-phase-name-here.md
└── ...

Research Output Requirements

• Ensure every research markdown report remains concise (<=150 lines) while covering all requested topics and citations.

Plan File Specification

• Every plan.md MUST ATTENTION start with YAML frontmatter: --- title: '{Brief title}' description: '{One sentence for card preview}' status: pending priority: P2 story_points: {1-21 modified fibonacci} effort: {sum of phases, e.g., 4h} branch: {current git branch} tags: [relevant, tags] created: {YYYY-MM-DD} ---
• Save the overview access point at {plan-dir}/plan.md. Keep it generic, under 80 lines, and list each implementation phase with status, progress, parallelization group, and links to phase files.
• For each phase, create {plan-dir}/phase-XX-phase-name-here.md containing the following sections in order:

- Context links (reference parent plan, dependencies, docs) - Parallelization Info (which phases can run concurrently, which must wait) - Overview (date, description, priority, implementation status, review status) - Key Insights - Requirements - Architecture - Related code files (MUST ATTENTION be exclusive to this phase - no overlap with other phases) - File Ownership (explicit list of files this phase owns/modifies) - Implementation Steps - Todo list - Success Criteria - Conflict Prevention (how this phase avoids conflicts with parallel phases) - Risk Assessment - Security Considerations - Next steps

Main plan.md must include:

• Dependency graph showing which phases can run in parallel
• Execution strategy (e.g., "Phases 1-3 parallel, then Phase 4")
• File ownership matrix (which phase owns which files)

IMPORTANT Task Planning Notes (MUST ATTENTION FOLLOW)

• Always plan and break work into many small todo tasks using TaskCreate
• Always add a final review todo task to verify work quality and identify fixes/enhancements
• MANDATORY FINAL TASKS: After creating all planning todo tasks, ALWAYS add these three final tasks:

1. Task: "Write test specifications for each phase" — Add ## Test Specifications with TC-{FEAT}-{NNN} IDs to every phase file. Use /tdd-spec if feature docs exist. Use Evidence: TBD for TDD-first mode. 2. Task: "Run /plan-validate" — Trigger /plan-validate skill to interview the user with critical questions and validate plan assumptions 3. Task: "Run /plan-review" — Trigger /plan-review skill to auto-review plan for validity, correctness, and best practices

Important Notes

IMPORTANT: Analyze the skills catalog and activate the skills that are needed for the task during the process. IMPORTANT: Ensure token efficiency while maintaining high quality. IMPORTANT: Sacrifice grammar for the sake of concision when writing reports. IMPORTANT: In reports, list any unresolved questions at the end, if any. IMPORTANT: Each phase MUST ATTENTION have exclusive file ownership - no file can be modified by multiple phases.

REMINDER — Planning-Only Command

DO NOT use EnterPlanMode tool. DO NOT start implementing. ALWAYS validate with /plan-review after plan creation. ASK user to confirm the plan before any implementation begins. ASK user decision questions with your recommendations when multiple approaches exist.

Next Steps (Standalone: MUST ATTENTION ask user via AskUserQuestion. Skip if inside workflow.)

MANDATORY IMPORTANT MUST ATTENTION — NO EXCEPTIONS: If this skill was called outside a workflow, you MUST ATTENTION use AskUserQuestion to present these options. Do NOT skip because the task seems "simple" or "obvious" — the user decides:

• "Proceed with full workflow (Recommended)" — I'll detect the best workflow to continue from here (plan created). This ensures review, validation, implementation, and testing steps aren't skipped.
• "/plan-review" — Auto-review plan for validity and best practices
• "/plan-validate" — Interview user to confirm plan decisions
• "Skip, continue manually" — user decides

If already inside a workflow, skip — the workflow handles sequencing.

Closing Reminders

• IMPORTANT MUST ATTENTION break work into small todo tasks using TaskCreate BEFORE starting
• IMPORTANT MUST ATTENTION search codebase for 3+ similar patterns before creating new code
• IMPORTANT MUST ATTENTION cite file:line evidence for every claim (confidence >80% to act)
• IMPORTANT MUST ATTENTION add a final review todo task to verify work quality
• IMPORTANT MUST ATTENTION include Test Specifications section and story_points in plan frontmatter
• IMPORTANT MUST ATTENTION search 3+ existing patterns and read code BEFORE any modification. Run graph trace when graph.db exists.
• IMPORTANT MUST ATTENTION include story_points and complexity in plan frontmatter. SP > 8 = split.
• IMPORTANT MUST ATTENTION include ## Test Specifications with TC IDs per phase. Call TaskList before creating new tasks.
• IMPORTANT MUST ATTENTION score complexity first. Score >=6 → decompose. Each phase: plan → implement → review → fix → verify. No skipping.