SuperAPI AI Directory

skill-refine

Coding & development

View repository

Updated 10-16 06:39

About this skill

Skill Refine

Convert feedback into plan-level patches. Do not edit the skill bundle directly.

Quick start

Agent: "refine plan skills/my-skill/.skill-plan.yaml"

Workflows

1. Load plan + reports

  1. Require existing .skill-plan.yaml path. If missing, stop and ask user for path.
  2. Read base plan YAML. Confirm it is parseable and has required planning fields.
  3. Locate *-skill-check-report.html and/or *-eval-report.html provided by user.
  4. Run Bash scripts/parse-report.sh <report-path> for each report.
  5. Validate parsed JSON against assets/report-schema.json. Show parse errors if malformed.
  6. Present feedback summary: total findings, top failures, affected plan areas.

2. Rank findings

  1. Read references/priority-rules.md and classify findings into P0-P3.
  2. Sort findings by tier, then impact within tier.
  3. Present ranked queue to user. Include counts per priority tier.

3. Propose plan patch operations

  1. For top-ranked finding, generate one plan patch proposal using references/change-proposal-format.md.
  2. Scope patches to plan fields only (e.g., intent, scope, mechanism, context_assets, workflow_notes, example_placed, target_path).
  3. Present options: [apply / skip / edit / stop].
  4. On apply, append operation to patch queue.
  5. On edit, accept user's modified patch operation and re-present once for confirmation.
  6. On skip, advance to next finding. On stop, persist queue and exit loop.

4. Build artifacts

  1. Write queued operations to <plan-dir>/.skill-plan.patch.yaml using assets/plan-patch-template.yaml shape.
  2. Apply operations to base plan in memory and produce <plan-dir>/.skill-plan.patched.yaml.
  3. Run Bash scripts/validate-plan.sh <plan-dir>/.skill-plan.patched.yaml to validate forge import contract.
  4. If validation fails, mark failing operations and ask user whether to edit/drop them.
  5. On validation pass, show concise plan diff summary (changed fields only).

5. Handoff

  1. Ask user to approve patched plan for forge input.
  2. On approval, instruct: /skill-forge <plan-dir>/.skill-plan.patched.yaml.
  3. On reject, keep patch artifact and return to step 10.

Guardrails

  • Human-in-the-loop always. Never auto-apply patches without explicit user approval.
  • Never edit target skill bundle files directly in this skill.
  • Keep patch operations granular (one concern per operation).
  • Preserve original .skill-plan.yaml; patched output goes to .skill-plan.patched.yaml.

Reference files

See priority-rules.md for finding severity tiers and ranking logic. See change-proposal-format.md for patch proposal template. See Example.md for end-to-end walkthrough. See report-schema.json for parsed report JSON schema. See plan-patch-template.yaml for patch artifact shape. See changelog-template.md for patch log entry format.