agentleFS
Sign inStart free
leosiminileosimini/keelokit/skills/coordinate/SKILL.md

coordinate

Skill2 starsChanged 10 days agoMITView on GitHub

Run several Keelokit sessions at once and keep them honest — one session per story with the shared agent brief, status in ten lines, follow-up rounds, and verifying landed work by its hash on origin/main. Use when the user says "coordinate", "run these stories in parallel sessions", "status of the sessions", "¿cómo van las sesiones?", "coordiná las sesiones", "estado de las sesiones", or when several sessions or worktrees are running.

22 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use addskill, apply-firmware 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.
  • ryanportfolioSkill

    addskill

    Create, import, update, or install repo or personal skills: runtime ownership, resources, discovery checks.

    Also in 6 projects that use this

  • ryanportfolioSkill

    apply-firmware

    Bring current folder (plain folder, existing repo, old or partial firmware copy) up to full Harness Firmware in place. Mirror external repo → adopt-repo.

    Also in 6 projects that use this

  • +10 more from ryanportfolio

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 21 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

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

    Wait with one until loop on one condition, never with repeated polling.

    line 36

“Common” is measured: kept by at least 6 in 10 of the 21 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

Coordinate — origin/main is the only truth

The brief every session follows is .keelokit/harness/agent-brief.md (SESS-1). Read it once.

First read [landing] policy in .keelokit/state.toml (absent = direct). With convoy, say in one line that coordination is /harbor:*'s, and only run and report python3 .keelokit/bin/status.py; the rest of this skill is for direct.

Start sessions

  • One session per story, from the same wave (the doctor guarantees their touches don't overlap). Stories that add migrations go one at a time (MIG-1).
  • The prompt is short: Read .keelokit/harness/agent-brief.md, then /keelokit:build-story <ID>. Pass paths, not contents; don't repeat the brief's rules.
  • Title each session with python3 .keelokit/bin/status.py title <ID> (create_session's title), and rename this one with title --role coordinator when the tool can. Exit 2 means the story or its wave is missing: fix the backlog, never type a title by hand. Branches stay the id in lowercase.
  • Respect the machine: at most two sessions running tests at once (pnpm verify takes a per-machine lock, so more only wait).

Every round (about every 15 minutes while sessions run)

  1. python3 .keelokit/bin/status.py: worktrees, branches with commits not on origin/main, the verify lock, the CI of main (a cancelled run is not green) and the backlog, in ten lines.
  2. For each session that says it finished: git fetch origin main && git log origin/main --oneline and find its commit (Story: <ID> trailer). A session that ends well does not message you; don't wait for a message, check the hash. "Done" without a commit on main is not done: send it back to land (rebase, pnpm verify, push) or land it yourself.
  3. CI of main red, or two runs in a row without a green: stop starting stories; python3 .keelokit/bin/ci.py jobs says which job and step; fix that first.
  4. Wait with one until loop on one condition, never with repeated polling.

Report to the user

Two lines per session at most: story, hash on main (or "not landed" and why), then what needs the user. The whole report fits in ten lines.

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