agentleFS
markmhendricksonmarkmhendrickson/neotoma/.claude/skills/run_feature_workflow/SKILL.md

run_feature_workflow

Skill38 starsChanged 2 months agoMITView on GitHub

Run Feature Workflow

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 8 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read foundation-config.yaml to get feature unit settings

    line 16
  2. Distinctive:

    Update spec/manifest if ambiguous

    line 33
  3. Very specific:

    Update code/tests/docs together

    line 34
  4. Distinctive:

    Follow architectural invariants (from repo-specific docs)

    line 35
  5. Distinctive:

    Verify coverage meets configured targets

    line 42
  6. Distinctive:

    Do NOT attempt to fix bugs inline

    line 46
  7. Distinctive:

    After bugfix completes, re-run the test suite

    line 48
  8. Distinctive:

    Only proceed to step 6 once all tests pass

    line 49
  9. Distinctive:

    Before producing the final summary:

    line 55
  10. Distinctive:

    Do NOT proceed to final summary until spec and implementation are aligned (either by updating spec or fixing implementation)

    line 65

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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
<!-- Source: .cursor/skills/run-feature-workflow/SKILL.md -->

Run Feature Workflow

feature_id = {{input:feature_id}}

Configuration is read from foundation-config.yaml.

  1. Load configuration:
    • Read foundation-config.yaml to get feature unit settings
    • Determine directory structure, manifest complexity, testing requirements
  2. Read authoritative documents:
    • foundation/development/feature_unit_workflow.md (workflow context)
    • {configured_directory}/in_progress/{{input:feature_id}}/{{input:feature_id}}_spec.md (or completed/ if completed)
    • {configured_directory}/in_progress/{{input:feature_id}}/manifest.yaml (or completed/ if completed)
    • foundation/development/templates/feature_unit_spec_template.md
    • Repository-specific execution instructions (if they exist)
    • Repository-specific architecture/module documentation (based on spec)
  3. Determine risk level:
    • low → skip plan
    • medium → produce brief structured plan
    • high/critical → produce detailed plan and wait for user approval
  4. Apply Feature Unit execution principles:
    • Update spec/manifest if ambiguous
    • Update code/tests/docs together
    • Follow architectural invariants (from repo-specific docs)
    • Preserve accessibility and i18n rules (if applicable)
  5. Run tests (based on configuration):

    • Unit tests (if testing.require_unit_tests: true)
    • Integration tests (if testing.require_integration_tests: true)
    • E2E tests (if testing.require_e2e_tests: true)
    • Verify coverage meets configured targets

    If any tests fail:

    • Immediately invoke Fix Feature Bug with feature_id={{input:feature_id}}
    • Do NOT attempt to fix bugs inline
    • Let the error protocol classify and correct
    • After bugfix completes, re-run the test suite
    • Only proceed to step 6 once all tests pass
  6. Apply error protocol (if repository has one) only if bugs were encountered and fixed via step 5.
  7. Spec-to-Implementation Diff Checkpoint (CRITICAL):

    Before producing the final summary:

    • Compare the implemented code against the spec
    • List any ways implementation intentionally diverged from spec (even if small)
    • For each divergence:
      • Either update the spec to reflect the change (if divergence is intentional and correct)
      • Or mark the divergence as a bug to be fixed (if divergence is unintentional)
    • Present this diff to the user:
      • "Implementation vs Spec Diff: [list of divergences]"
      • "For each divergence, should I update the spec or fix the implementation? (spec/implementation/list)"
    • Do NOT proceed to final summary until spec and implementation are aligned (either by updating spec or fixing implementation)
  8. Produce a final change summary:
    • feature_id
    • files changed
    • risk
    • error_class (if bugfix)
    • tests run
    • coverage achieved

Inputs

  • feature_id (string): The feature_id to work on.

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.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

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.