agentleFS
Sign inStart free
rorerore/Pallium/.claude/skills/agent-workflow/SKILL.md

agent-workflow

Skill2 starsChanged 12 days agoMITView on GitHub

Use for implementation plans or repository changes, explicit Agent Workflow requests, and existing workflow task actions. Skip standalone read-only review, explanation, diagnosis, comparison, or inspection.

20 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use agent-agent, agent-analyze-code-quality and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 15 edited versions we read keep these.

  • 2

    Distinctive

    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. Distinctive:

    Read only the file for the mode that applies.

    line 16
  2. Distinctive:

    Pause on assumption failure: Stop and update the assumption + affected fields.

    line 24
  3. Very specific:

    Report upstream defects: If product/skill behavior fails, contradicts guidance, or forces a consequential guess, load …

    line 26
  4. Very specific:

    When ambiguous: repo's agent-workflow.yaml → repo's AGENTS.md/CLAUDE.md → mode-specific skill file → these principles → ask the developer.

    line 30

“Common” is measured: kept by at least 6 in 10 of the 15 edited versions of this file we could read. “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

agent-workflow

Drive engineering tasks through a small set of checkpoints. The harness records what you do in a structured Work Record so the next agent or reviewer can pick up where you stopped. Plan quality, discovery sufficiency, implementation correctness, evidence adequacy — those stay reviewer judgments; you produce the artifacts that make them possible.

Pick a mode

  1. agent-workflow.yaml exists in the repo root → read operating-mode.md. Everyday work.
  2. User asked you to set up agent-workflow on this repo → read bootstrap-mode.md. One-shot install.
  3. Neither → this skill is not relevant.

Read only the file for the mode that applies.

Principles

  • Gate request scope, then applicability. Standalone read-only analysis stays out; implementation plans, repository changes, explicit use, and existing-task actions enter. For in-scope work, only a whole-change applicability exemption skips the Work Record.
  • Shape is derived from (Risk, Complexity). (Routine, Simple) → compact. Anything else → expanded. Checker enforces via workrecord.shape_matches_classification.
  • Stay within recorded scope. Material scope expansion returns the task to planning.
  • Update State at transitions. Ready to implement → Ready for review only when verification passes.
  • Pause on assumption failure. Stop and update the assumption + affected fields. Do not continue under stale assumptions.
  • Marker block is the structured surface. Prose around it is human notes. Don't put structured state outside the markers; don't put free prose inside them.
  • Report upstream defects. If product/skill behavior fails, contradicts guidance, or forces a consequential guess, load templates/skill-feedback.md. Review the Result is the final catch.

Decision priority

When ambiguous: repo's agent-workflow.yaml → repo's AGENTS.md/CLAUDE.md → mode-specific skill file → these principles → ask the developer. Do not invent rules the config does not state.

Resources (load on demand)

  • operating-mode.md — when entering operating mode.
  • bootstrap-mode.md — when entering bootstrap mode.
  • templates/checkpoints/*.md — per-checkpoint, load only when entering that checkpoint.
  • templates/skill-feedback.md — load only when a trigger from the Skill-feedback check fired.
  • The repo's agent-workflow.yaml — once on pickup, to learn the configured taskPath.

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.

Check your setup, free