Agent skill
planning
Research-backed implementation planning skill. This skill should be used when the /plan command is executed, when creating implementation plans, roadmaps, or architectural designs, or when tasks have high uncertainty requiring research before coding. Transforms user requests into execution-ready plans (PRP) using R.P.I.V workflow (Research → Plan → Implement → Validate).
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/planning-grupous-gpus-3
SKILL.md
Planning Skill
Research-first planning methodology that eliminates unknowns before implementation.
Core Principle: Context Density > Brevity | Research First > Implementation | Planning > Coding
Activation Triggers
This skill is mandatory when:
- User executes
/plancommand - Building plans, roadmaps, or architecture for new features/systems
- High uncertainty or risk of hallucination without research
- Multi-step execution requiring task decomposition
- Third-party integrations (APIs, frameworks, infrastructure)
Skip for: Simple Q&A, pure copywriting, tasks solvable from provided context.
R.P.I.V Workflow
Phase 0: RESEARCH (Always First)
Eliminate unknowns and lock in best-practice approach.
Priority order:
- Search codebase for patterns, conventions
- Query Context7 for official docs
- Tavily web search for best practices
- Sequential Thinking for complex decisions
Required outputs:
- Findings Table:
| # | Finding | Confidence (1-5) | Source | Impact | - Knowledge Gaps: What remains unknown
- Assumptions to Validate: Explicit assumptions needing confirmation
- Edge Cases: Minimum 5 for L4+ complexity
See
references/mcp-usage.mdfor detailed MCP tool guidance.
Phase 1: PLAN (Before Implementation)
Convert research into execution runbook with atomic tasks.
Each task must have:
- Clear action verb + specific target
- Validation command
- Rollback steps
- Dependencies mapped
- Mark
⚡ PARALLEL-SAFEwhen independent
Output: docs/PLAN-{task-slug}.md
See
references/plan-template.mdfor complete template structure.
Phase 2: IMPLEMENT (Only If Requested)
Execute per atomic tasks with validation gates.
Pattern: Implement → Validate → Commit (or Rollback)
Quality gates after each task:
bun run check(TypeScript)bun run format(Prettier)bun test(Unit tests)
Phase 3: VALIDATE (Always)
- Build: zero errors
- Lint: zero warnings
- Tests: all passing
- Consult specialists if security or schema changes involved
Operating Modes
| Mode | Behavior | Output |
|---|---|---|
CONSERVATIVE (default for /plan) |
Research + plan only, no code | docs/PLAN-{slug}.md |
| PROACTIVE (for implementation) | Research → plan → implement | Code + plan file |
Complexity Levels
| Level | Indicators | Research Depth |
|---|---|---|
| L1-L2 | Bug fix, single function | Repo-only |
| L3-L5 | Feature, multi-file | Docs + repo |
| L6-L8 | Architecture, integration | Deep research |
| L9-L10 | Migrations, multi-service | Comprehensive |
See
references/complexity-guide.mdfor detailed classification.
Anti-Patterns
| Bad | Good |
|---|---|
| "Implement auth" | Research → Search codebase → Query docs → Then implement |
| Skip research | ALWAYS research first, even for "simple" tasks |
| Guess file paths | Search and verify paths before referencing |
| Speculate about code | Read files before making claims |
Quick Reference
R.P.I.V: RESEARCH → PLAN → IMPLEMENT → VALIDATE
GOLDEN RULES:
✓ RESEARCH FIRST — never implement blind
✓ Be EXPLICIT — agent follows literally
✓ Explain WHY — enables generalization
✓ ATOMIC TASKS — small, validated, rollback-ready
✓ REFLECT AFTER TOOLS — think before next action
Post-Planning Message
After creating the plan file:
✅ Plan created: docs/PLAN-{slug}.md
Next steps:
- Review the plan
- Run `/implement` to start implementation
- Or modify plan manually
Resources
references/plan-template.md— Complete PRP template structurereferences/complexity-guide.md— Task complexity classificationreferences/mcp-usage.md— When and how to use MCP tools
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?