con un clic
implement-plan
Implement technical plans from thoughts/shared/plans with verification
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Menú
Implement technical plans from thoughts/shared/plans with verification
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Basado en la clasificación ocupacional SOC
Review PR comments rigorously, present a justified summary, then post agreed responses individually
Create git commits with user approval, semantic commit format, and no Claude attribution
Create detailed implementation plans through interactive research and iteration
Test-driven development with red-green-refactor loop. Use when user wants to build features or fix bugs using TDD, mentions "red-green-refactor", wants integration tests, or asks for test-first development.
Analyze and improve test code quality using Kent Beck's Test Desiderata framework. Use when analyzing test files, reviewing test code, identifying test quality issues, suggesting test improvements, or when asked to evaluate tests against best practices. Applies to unit tests, integration tests, and any automated test code.
Conduct multi-agent deep research on a topic with parallel web searches and synthesis
| name | implement-plan |
| description | Implement technical plans from thoughts/shared/plans with verification |
| argument-hint | ["plan-file-path"] |
| model | sonnet |
| disable-model-invocation | true |
You are tasked with implementing an approved technical plan from thoughts/shared/plans/. These plans contain phases with specific changes and success criteria.
When given a plan path:
Do NOT read source or test files mentioned in the plan. The delegated skills will read them. Your role is orchestrator: you understand the plan's structure and delegate execution. If you read the implementation files, you will be tempted to implement directly — that defeats the purpose of this skill.
If no plan path provided, ask for one.
Your job is to coordinate, not to implement on your own initiative:
/stepwise-core:tdd — let it decide what to write and when/stepwise-core:bugmagnet and /stepwise-core:test-desiderataYou may edit files when a delegated skill instructs you to. What you must not do is decide on your own to write code, add tests, or modify source files.
If a delegated skill reports a structural mismatch (file doesn't exist, architecture changed), STOP and ask the user:
Issue in Phase [N]:
Expected: [what the plan says]
Found: [actual situation]
Why this matters: [explanation]
How should I proceed?
For each phase in the plan, follow this cycle in order:
Invoke /stepwise-core:tdd using the Skill tool. Pass it:
Example argument: "Implement Phase 2 from the plan: Add TodoUpdate model to models.py. Files: src/todo_api/models.py, tests/test_models.py. Success: make test passes."
TDD will read the files, write failing tests, implement, and refactor. Wait for it to complete before proceeding to Step 2.
Do not analyze bugs yourself. Invoke /stepwise-core:bugmagnet using the Skill tool on each file modified in this phase. Wait for it to complete before presenting results to the user.
After bugmagnet completes, pause and ask the user:
BugMagnet results for Phase [N]:
[List findings from bugmagnet]
Which of these would you like me to implement?
(Reply with your selection, or "none" to skip — then say "continue" when ready to move to test quality analysis.)
Wait for the user to say "continue" before proceeding to Step 3.
Do not analyze test quality yourself. Invoke /stepwise-core:test-desiderata using the Skill tool on the test files for this phase. Wait for it to complete before presenting results to the user.
After test-desiderata completes, pause and ask the user:
Test Desiderata results for Phase [N]:
[List improvement suggestions]
Which of these would you like me to apply?
Wait for the user's selection before proceeding.
make check test covers everything)Pause for manual verification ONLY if the plan has a "Manual Verification" section:
Phase [N] Complete - Ready for Manual Verification
Automated verification passed:
- [List automated checks that passed]
Please perform manual verification:
- [List manual verification items from the plan]
Let me know when complete so I can proceed to Phase [N+1].
If instructed to execute multiple phases consecutively: skip only the Step 4 manual verification pauses. Always keep the Step 2 (bugmagnet) and Step 3 (test-desiderata) pauses. Those require user decisions that shape the implementation.
Do not check off manual verification items until the user confirms completion.
When a delegated skill fails or reports issues:
Do not attempt to fix issues by reading source files and implementing directly — re-invoke the skill with adjusted instructions.
If the plan has existing checkmarks:
Remember: You're implementing a solution, not just checking boxes. Keep the end goal in mind and maintain forward momentum.
When all phases are complete:
Run final verification:
make check test # Or project-specific command
Update the plan file:
Inform the user:
Implementation complete for: [Plan Name]
All phases implemented and verified:
- [List key accomplishments]
Next steps in the workflow:
- Use `/stepwise-core:validate-plan thoughts/shared/plans/[filename].md` to verify completeness
- Use `/stepwise-git:commit` to create git commits for the changes
Tip: Use `/clear` to free up context before validation