agentleFS
Sign inStart free
thepictishbeastthepictishbeast/claude-tools/skills/fleet/SKILL.md

fleet

Skill1 starsChanged 6 months agoMITView on GitHub

Orchestrate the PlausiDen agent fleet (Conductor + worker swarm). Validate a multi-agent plan's scope-disjointness, dispatch a run, run the verify/burn gates, inspect live status or an individual worker, and engage/clear the global STOP. Use when the user wants to run, plan, gate, inspect, or stop the fleet, or open the fleet console. Heavy ops defer to the `fleet` MCP server.

18 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use learn and triage.

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.
  • genomewalkerSkill

    learn

    Learn from web sources, generate SKILL.md, and store key patterns in soul memory. Trigger keywords: learn, skill-gen, web-learn, documentation, tutorial.

    Also in 3 projects that use this

  • Google CloudSkill

    triage

    Gather context on a change or bug and plan the implementation.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is 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.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never run the fleet against a live / auto-deployed tree: Use safe-test mode (isolated repo copy).

    line 27
  2. Very specific:

    Don't dispatch a live run without fleet_validate_plan passing first.

    line 35
  3. Distinctive:

    Don't bypass the STOP / safe-test guards to "go faster".

    line 36

“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

/fleet — agent-fleet orchestration (MCP-backed)

The fleet's control surface ships as the fleet MCP server (fleet mcp, registered with Claude). Prefer the MCP tools over shelling fleet — they are typed, return structured JSON, and lazy-load their schemas (you only pay context for tools you actually call). Reach for these before bash.

MCP tools (the Conductor's control surface)

ToolWhatKind
fleet_validate_planRAIL 2 dry check: parse a plan, report workers + whether scope-claims are disjoint (disjoint:false is a refusal verdict, not an error)read-only
fleet_runRun a plan. Default dry-run; live:true enforces RAIL 2 then dispatches workers detached — poll fleet_statusgated dispatch
fleet_burnQuality-token-burn: run live at high concurrency on real backlog (429-retry + result-accept gate)dispatch
fleet_gateRAIL 3 result-accept gate: check the working tree against the plan's claims (strays = changes outside all claims)read-only
fleet_statusBus snapshot: worker messages, kill-switch state, latest result-accept verdictread-only
fleet_stop / fleet_resumeEngage / clear the global kill switch (STOP)control
fleet_worker_statusOne worker's live state — status, scope-claim, attempt, last heartbeat/finding, drift flagsread-only
fleet_worker_memoryTail a worker's own running notes (memory.md)read-only
fleet_worker_diffA worker's changes within its scope-claim (paths + git diff --stat)read-only

Safety (load-bearing — see the 2026-06-03 incident)

  • Never run the fleet against a live / auto-deployed tree. Use safe-test mode (isolated repo copy). Workers once edited the live ProsperityClub tree and broke dev.plausiden.com.
  • fleet_validate_plan (RAIL 2, scope-disjointness) and fleet_gate (RAIL 3, result-accept) are hard gates, not advisory. Trust the deterministic gate over any worker's self-review.
  • fleet_stop is the kill switch — engaging it refuses all further spawns until fleet_resume.

Console

The live command-center is fleet-console on :8848 (SSE-driven; dev4.plausiden.com is the deployed static dashboard). It exposes message / upload / command / stop-resume. Spawn + per-agent settings are intentionally gated stubs pending safe-test wiring.

Don't

  • Don't dispatch a live run without fleet_validate_plan passing first.
  • Don't bypass the STOP / safe-test guards to "go faster".

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