agentleFS
cofincofin/flow/skills/flow-completion/SKILL.md

flow-completion

Skill19 starsChanged 5 days agoApache-2.0View on GitHub

Use when reviewing (/flow:review), finishing (/flow:finish), archiving (/flow:archive), cleaning up (/flow:cleanup), reverting (/flow:revert), or validating (/flow:validate) Flow work.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    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. Very specific:

    Use for review (../flow/references/review.md), finish (../flow/references/finish.md), archive (../flow/references/archive.md), cleanup

    line 13
  2. Very specific:

    Verify branch-local completion eligibility: every chapter (chapters/<ch_id>/spec.md) and task worksheet is completed (or deferred [-]) with …

    line 33
  3. Very specific:

    Run full verification gates (knowledge/workflow.md), confirm flow-wide review gates passed, evaluate the ADR Filter …

    line 34
  4. Very specific:

    Run groundskeeping and debloat across touched files (/flow:cleanup), revert designated task, chapter, phase, or flow commits while …

    line 39
  5. Very specific:

    Never stamp last_reviewed_commit or close a chapter/flow until both code/tests AND present-tense Markdown/OKF/user docs (spec.md, …

    line 44
  6. Very specific:

    Never delete a completed spec directory during /flow:archive until its durable discoveries and patterns across all chapters have been …

    line 45

≈ “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

Flow Completion

<!-- lifecycle-ownership: owner=flow-completion; operations=review,finish,archive,cleanup,revert,validate -->

Trigger

Use for review (../flow/references/review.md), finish (../flow/references/finish.md), archive (../flow/references/archive.md), cleanup (../flow/references/cleanup.md), revert (../flow/references/revert.md), and validate (../flow/references/validate.md).

<!-- quality-completion-policy: start -->
contract: quality-completion-v1
authority: skills/flow/references/review.md
finish_gates: [verification, code_review, quality_review, finish]
archive_gates: [archive_candidate, verification, code_review, quality_review, knowledge_contraction, archive]
runtime_dependency: agent_file_tools_only
<!-- quality-completion-policy: end -->

Workflow

  1. Zero-Context Chapter or Flow Correctness, Markdown-Sync, & Quality Review (/flow:review [flow_id] [chapter_id]):
    • Compute the exact base_commit..head_commit range for the target chapter or full flow and inspect the changed code, test, spec, OKF, and doc surface.
    • Dispatch isolated zero-context reviewers (run_workflow parallel() with type_name="self" on Google Antigravity, checking git status --porcelain afterward; isolated @code-reviewer, @quality-reviewer, and matching .agents/agents/<domain>-reviewer subagents elsewhere) passing only disk paths (spec.md, chapters/index.md, chapters/<ch_id>/spec.md, tasks/*.md, .agents/skills/, and knowledge/ chapters) under the Zero-Context Isolation Rules.
    • Evaluate Spec Conformance, Present-Tense Markdown, Full-Plan Roadmap, & Doc-Sync Conformance, Repository Standards, Deep-Module Seams, & Baseline Structural Code Smells, and Mandatory Quality, Zero-History Doc Hygiene, & Debloat Review (quality-review-v1).
    • Adjudicate findings (ADOPT vs. CONTEST vs. ESCALATE), remediate adopted code or Markdown-drift findings, re-verify, and stamp last_reviewed_commit: <REVIEWED_HEAD_SHA> on the reviewed chapter/flow spec.md.
  2. Finish Flow (/flow:finish):
    • Verify branch-local completion eligibility: every chapter (chapters/<ch_id>/spec.md) and task worksheet is completed (or deferred [-]) with a valid last_reviewed_commit covering all chapter task commits and [N/N files documented] in ## 3. Documentation & OKF Knowledge Plan.
    • Run full verification gates (knowledge/workflow.md), confirm flow-wide review gates passed, evaluate the ADR Filter (knowledge/decisions/adr-NNN-<slug>.md), mark root spec.md state: completed (current_chapter: null, current_task: null), and present the structured completion options (Archive flow now, Archive and open/update PR, Merge locally and archive, Discard work).
  3. Archive & Contract Flow (/flow:archive):
    • Execute the subagent-sharded /flow:learn contraction pass (1 subagent per domain when >= 2 domains) across all chapters, OKF knowledge namespaces, and .agents/skills/** before deletion.
    • Enforce strict write ordering: write knowledge/** + .agents/skills/** first, prepend the archive entry to .agents/bundles/log.md second, and delete .agents/bundles/specs/<flow_id>/ (including chapters/) last.
  4. Cleanup, Revert, or Validate (/flow:cleanup, /flow:revert, /flow:validate):
    • Run groundskeeping and debloat across touched files (/flow:cleanup), revert designated task, chapter, phase, or flow commits while resetting chapter/root roadmaps in present tense (/flow:revert), or run the OKF v0.2 multi-chapter bundle, hook, skill, and spec validator (/flow:validate).

Guardrails

  • Reviewers are strictly read-only isolated agents and receive zero conversational history, rationale, or leading hints.
  • Never stamp last_reviewed_commit or close a chapter/flow until both code/tests AND present-tense Markdown/OKF/user docs (spec.md, chapters/index.md, knowledge/**, docs/**) are synchronized and verified free of historical prose.
  • Never delete a completed spec directory during /flow:archive until its durable discoveries and patterns across all chapters have been synthesized into .agents/bundles/knowledge/** and logged in .agents/bundles/log.md.
  • Preserve Git history and never create or mutate Git tags.

Output

Return the review adjudication summary (ADOPT / CONTEST / ESCALATE), verification command exit codes, updated spec.md state or .agents/bundles/knowledge/ contraction report, and next action.

Validation

Confirm all required review and verification gates passed with exit code 0, spec.md and tasks/*.md states match Git history, and .agents/bundles/index.md and log.md are updated cleanly.

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.