agentleFS
me-cedricme-cedric/agent-compass/skills/spec-flow/SKILL.md

spec-flow

Skill3 starsChanged 27 days agoMITView on GitHub

Write, validate and archive OpenSpec specs (konvoy-delivery schema); map screens, convert docs, draft specs from code, triage drift. Use before code for a feature or unclear task.

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

25 rules. About 0 are common, 9 are distinctive, and 16 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈9

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 16

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the smallest spec that prevents a misunderstanding.

    line 11
  2. Very specific:

    Read status before each step.

    line 41
  3. Very specific:

    Run validate before you report done.

    line 42
  4. Very specific:

    Run ac check --docs before you report done.

    line 44
  5. Distinctive:

    Use this to start a project from input documents: a brief, a tender, meeting notes.

    line 49
  6. Very specific:

    Run ac init --from-specs <dir>.

    line 51
  7. Very specific:

    Run grill on the inbox. It writes one RAID file for each open question.

    line 53
  8. Very specific:

    Write {{OPENSPEC_DIR}}/constitution.md if it does not exist.

    line 54
  9. Distinctive:

    Write one change for each feature: the proposal first, then the quartet for each capability.

    line 55
  10. Distinctive:

    Keep the inbox files. They are the source of each requirement.

    line 56
  11. Very specific:

    Use this for a project with spec-kit (.specify/, specs/<feature>/), a root specs/ folder or a single-file question list.

    line 60
  12. Very specific:

    Run ac migrate --docs --dry.

    line 62
  13. Very specific:

    Run ac migrate --docs only when the user says yes.

    line 63
  14. Very specific:

    Do each manual item with the user.

    line 64
  15. Very specific:

    Run ac check --docs. Report each openspec validate failure.

    line 66
  16. Very specific:

    Read references/delta-section.md when you append the delta section to a spec.

    line 70
  17. Distinctive:

    Use this for a codebase that has no specs.

    line 92
  18. Distinctive:

    Read the entry points, routes, jobs, migrations, schemas and tests.

    line 95
  19. Very specific:

    Put this line under the title of each draft: > Inferred from code, not reviewed.

    line 96
  20. Distinctive:

    Add the sections "Open questions" and "Not covered" to each draft.

    line 97
  21. Distinctive:

    Never invent a requirement.

    line 98
  22. Distinctive:

    Never overwrite a spec. Write the draft beside it and name the conflict.

    line 99
  23. Very specific:

    Read references/openspec.md for the root rule, the config.yaml rules and a checklist gate.

    line 104
  24. Very specific:

    Read references/drift-triage.md when a spec changes after tickets exist.

    line 108
  25. Very specific:

    Read references/convert-documents.md when a task needs the content of a .docx, .pptx, .xlsx, .odt, .rtf, .epub or .pdf file.

    line 112

≈ “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

spec-flow

Use the smallest spec that prevents a misunderstanding. A tiny edit needs only an inline brief: goal, non-goal, acceptance, validation.

Layout

OpenSpec lives in {{OPENSPEC_DIR}}. The schema is konvoy-delivery.

{{OPENSPEC_DIR}}/constitution.md
{{OPENSPEC_DIR}}/changes/<id>/proposal.md
{{OPENSPEC_DIR}}/changes/<id>/specs/<cap>/{spec,plan,checklist,tasks}.md
{{OPENSPEC_DIR}}/changes/archive/

Flow

  1. Proposal: write what and why. Do not name a stack. Start with openspec new change <id>.
  2. Clarify: ask until no high-impact [NEEDS CLARIFICATION] remains. Use grill for a long interview.
  3. Specs: fill the headings of the schema template. Then append the delta section. See references/delta-section.md.
  4. Plan and tasks: write ordered tasks, tests first. Mark parallel-safe tasks with [P].
  5. Apply: implement only after the spec, plan and tasks are stable.
  6. Validate, then archive.
openspec status --change <id> --json
openspec instructions apply --change <id> --json
openspec validate --all --strict --json
openspec archive <id>

Read status before each step. Do not guess which artifact is ready. Run validate before you report done. Run archive only when the user asks.

Run ac check --docs before you report done. It checks the RAID files, the ADRs and the screen tags. It also runs openspec validate when the CLI is installed.

Init from specs

Use this to start a project from input documents: a brief, a tender, meeting notes.

  1. Run ac init --from-specs <dir>. It creates the layout and copies the documents into {{OPENSPEC_DIR}}/inbox/<dir name>/.
  2. Convert each document that is not Markdown. See references/convert-documents.md.
  3. Run grill on the inbox. It writes one RAID file for each open question.
  4. Write {{OPENSPEC_DIR}}/constitution.md if it does not exist. Ask for the governing rules. Do not invent them.
  5. Write one change for each feature: the proposal first, then the quartet for each capability.
  6. Keep the inbox files. They are the source of each requirement.

Migrate specs

Use this for a project with spec-kit (.specify/, specs/<feature>/), a root specs/ folder or a single-file question list.

  1. Run ac migrate --docs --dry. Show the plan to the user.
  2. Run ac migrate --docs only when the user says yes. Each source moves under migrated/. Nothing is deleted.
  3. Do each manual item with the user.
  4. Append the delta section to each migrated spec.md. See references/delta-section.md.
  5. Run ac check --docs. Report each openspec validate failure. Do not hide it.

Delta section

Read references/delta-section.md when you append the delta section to a spec.

Screens

The registry is {{DESIGNS_DIR}}/screens.json. It has one entry per screen:

{ "version": 1, "screens": [
  { "id": "SCR-014", "aliases": ["login"], "route": "/login",
    "figma": { "fileKey": "<key>", "nodeId": "12:345" },
    "stories": ["Auth/LoginForm"], "specRefs": ["auth/spec.md#REQ-003"] } ] }
  • Tag each requirement of a screen with the line Screen: SCR-014. An alias is also correct: Screen: login.
  • Match a name from the user against aliases first.
  • An alias belongs to one screen only.
  • Take the next free number for a new screen. Never reuse an id.
  • ac check fails on a tag that is not an id or an alias in the registry.
  • ui-conformance reads this registry.

Specs from code

Use this for a codebase that has no specs.

  1. Group the code into capabilities. Draft the eight largest first. List the others under "Not covered".
  2. Read the entry points, routes, jobs, migrations, schemas and tests. Use the code graph first when the project has one.
  3. Put this line under the title of each draft: > Inferred from code, not reviewed. Drafted from <commit sha>.
  4. Add the sections "Open questions" and "Not covered" to each draft.
  5. Never invent a requirement. Never present a likely defect as intended behavior.
  6. Never overwrite a spec. Write the draft beside it and name the conflict.
  7. Hand the open questions to grill.

OpenSpec CLI

Read references/openspec.md for the root rule, the config.yaml rules and a checklist gate.

Drift triage

Read references/drift-triage.md when a spec changes after tickets exist.

Convert documents

Read references/convert-documents.md when a task needs the content of a .docx, .pptx, .xlsx, .odt, .rtf, .epub or .pdf file.

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.