agentleFS
Marwan-Ahmed-devMarwan-Ahmed-dev/vibecode-pro-max/.agents/skills/vc-generate-plan/SKILL.md

vc:generate-plan

Skill0 starsChanged 3 months agoMITView on GitHub

Create or update implementation plans in the repo's SIMPLE or COMPLEX format. Use when turning an idea, PRD, or approved direction into a saved plan artifact.

We split this skill into the rules it gives an agent

11 rules. About 0 are common, 5 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill to produce the authoritative implementation plan artifact set for the kit's work.

    line 11
  2. Very specific:

    Read references/generate-plan.md for the full plan contract.

    line 24
  3. Very specific:

    Run date +%d-%m-%y before choosing the filename.

    line 25
  4. Very specific:

    If complexity is not obvious, ask whether the plan is SIMPLE or COMPLEX.

    line 26
  5. Very specific:

    Read process/context/all-context.md when present to choose relevant context docs.

    line 28
  6. Distinctive:

    Keep resume/dependency notes Markdown-structured for now.

    line 32
  7. Distinctive:

    If the work is a large multi-phase program, create or update a feature folder plan set:

    line 33
  8. Very specific:

    Prefer process/features/{feature}/active/ when the topic maps to an existing feature folder.

    line 46
  9. Very specific:

    Keep phase status honest: code-only completion is CODE DONE, not VERIFIED.

    line 47
  10. Distinctive:

    Treat validation failures as blockers before presenting the plan as ready.

    line 50
  11. Distinctive:

    Do not hide a large program inside one giant plan if execution will actually happen phase by phase.

    line 52

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Generate Plan

Use this skill to produce the authoritative implementation plan artifact set for the kit's work.

This skill is the canonical planning contract for the repo. Planning discipline previously spread across vc:plan now belongs here plus the plan-agent prompt.

Normal output is one plan file.

For large multi-phase programs, this skill instead defines how to create an umbrella plan plus phase-plan set under one feature folder. See process/development-protocols/phase-programs.md.

Optional input: a feature idea plus simple or complex when the user already knows the intended depth.

Workflow

  1. Read references/generate-plan.md for the full plan contract.
  2. Run date +%d-%m-%y before choosing the filename.
  3. If complexity is not obvious, ask whether the plan is SIMPLE or COMPLEX.
  4. Save the plan to process/general-plans/active/ unless the work belongs to an existing process/features/{feature}/active/ folder.
  5. Read process/context/all-context.md when present to choose relevant context docs.
  6. For complex plans, read process/context/planning/example-complex-prd.md before writing.
  7. Include automated and manual verification gates from process/context/tests/all-tests.md.
  8. For new or newly touched direct *_PLAN_*.md plans, include explicit sections for Touchpoints, Public Contracts, Blast Radius, Verification Evidence, and Resume and Execution Handoff.
  9. Keep resume/dependency notes Markdown-structured for now; do not invent a second machine-only schema.
  10. If the work is a large multi-phase program, create or update a feature folder plan set:
  • one umbrella/orchestration plan
  • one direct plan file per phase
  • one durable report destination per phase
  1. Validate the generated artifact:
node .claude/skills/vc-generate-plan/scripts/validate-plan-artifact.mjs <plan-path>

Important Rules

  • For standard work, create exactly one plan file.
  • For a phase program, create one umbrella plan plus one direct plan file per phase.
  • Prefer process/features/{feature}/active/ when the topic maps to an existing feature folder.
  • Keep phase status honest: code-only completion is CODE DONE, not VERIFIED.
  • Make execution trust explicit inside the plan: what code or data can change, what contracts are exposed, what proof is required, and how EXECUTE should resume after compaction.
  • End with the next instruction for RIPER-5 or Cursor Plan mode.
  • Treat validation failures as blockers before presenting the plan as ready.
  • Fold red-team questions, dependency mapping, verification gates, and ambiguity checks into the generated plan itself instead of relying on a parallel plan-owner workflow.
  • Do not hide a large program inside one giant plan if execution will actually happen phase by phase.
  • Preserve the older complex-plan behavior by keeping pre-phase research and proof gates inside each phase plan; the new protocol changes the artifact shape, not the rigor.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.