agentleFS
itqanlabitqanlab/agent-toolkit/.agents/skills/plane-fleet/SKILL.md

plane-fleet

Skill1 starsChanged 33 days agoMITView on GitHub

Run work from a Plane board with teams of agent roles, one section per kind of work. Triggers: 'plane fleet', 'fleet run', 'run the board', 'work the tickets', 'work PROJ-12', 'test PR 212', 'plan this', 'write a ticket for this', 'ship the queue on my own'. Engineering: a PM claims tickets oldest first, engineers each take one ticket into their own git worktree and environment slot and open a PR, testers check every PR blind in the browser and add tests, and the PM merges approved PRs in groups, confirms CI and deploy, and moves tickets to Done. Core: a planner turns requests into right-sized tickets, a board specialist files and reshapes them, a scanner does read-only counting. Roles can run on other coding CLIs too, picked per spawn by task kind and remaining budget, with real cost and time measured per ticket. Needs a Plane MCP server and the GitHub CLI.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    All paths below are relative to this folder, the one that holds this file.

    line 20
  2. Distinctive:

    Read the role file and follow it exactly.

    line 47

≈ “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

plane-fleet

Teams of agent roles that work a Plane board the way small teams do, one section per kind of work. Nobody waits for anybody: an engineer opens a PR and moves on, a tester picks up any open PR at once, the PM lands what is approved and green.

All paths below are relative to this folder, the one that holds this file.

Sections

SectionWhat it doesFolder
coreruns the fleet: the PM, the planner, the board specialist, the scanner, solo mode, the fleet command, the shared referencescore/
engineeringships code: an engineer per ticket, a tester per PR, the merge queueengineering/
socialposters, carousels and reels (planned: its roles are built and tested before they run)social/

Each section's README.md says what it holds. Each role is a folder with the same parts: role.md (who it is, its limits, its report, and an index of its skills) and skills/ (one file per thing it does, read only when that step comes).

Pick the role, then read its file

The user asksRoleFile
fleet run, run the board, work the tickets, keep the fleet goingPMcore/pm/role.md
plan this, break this down, size or regroup ticketsplannercore/planner/role.md
write a ticket for this, turn notes into ticketsplanner, writingcore/planner/skills/write-ticket.md
file this plan, merge these tickets, board reportboard specialistcore/board/role.md
count, triage, read these logsscannercore/scanner/role.md
work one ticket to a PRengineerengineering/engineer/role.md
test one open PRtesterengineering/tester/role.md
ship the queue alone, one ticket at a time, no teamsolocore/solo/role.md

Read the role file and follow it exactly. It names its skill files and when to read each.

Agents

The plugin's agents/ folder holds one short card per role, grouped by section, each carrying only the tools its role uses: core/ holds fleet-planner, fleet-board and fleet-scanner; engineering/ holds fleet-engineer and fleet-tester. A card points its agent at the role file. The PM spawns them as sub-sessions, one per ticket or PR. With no sub-sessions, run solo mode, or one engineer or tester role by hand. Install notes are in README.md.

Shared references

In the core section's references folder:

  • config-schema.md — the project config and the private per-user override
  • plane-api.md — the Plane MCP calls and what a board read costs
  • usage-budget.md — spending a usage window on purpose and stopping safely
  • executors.md and executors.json — which program runs which role, and prices
  • non-claude-adapter.md — what a role running on another coding CLI must know

In the engineering section's references folder:

  • standing-brief.md — the boundaries every engineer and tester keeps
  • ticket-rules.md — when to file a ticket, and how big one may be
  • brief-checklist.md — defect classes that caused repeat rework; the PM copies the ones that apply into each brief
  • notify.md — telling the client, with a screenshot

The fleet command

scripts/fleet runs lanes on other coding CLIs and holds the shared tools:

scripts/fleet config --cwd <repo>          # the project config, merged
scripts/fleet status                       # machine gate, budget, slots, lanes, open PRs, trunk CI
scripts/fleet resources                    # memory, swap, CPU and disk on any OS, the biggest processes, the fixes
scripts/fleet cache-server env --cwd <repo> # one Nx cache for every worktree (eval its output)
scripts/fleet slot-template clone A --cwd <repo>   # slot A's DB from the migrated, seeded template; exit 3 = run the recipe
scripts/fleet gate-stamp check --pr 212    # a group PR may merge on the local gate stamp
scripts/fleet watch PROJ-12-ci -- gh run watch 123   # one background watcher per name; a second start is refused
scripts/fleet pick --role engineer --kind api --cwd <repo>
scripts/fleet cost --since 2026-10-01      # real spend and time per merged ticket
scripts/fleet cost --ticket PROJ-12        # one ticket's sessions in order
scripts/fleet retro --since <run start> --cwd <repo>   # numbered findings on the run: cause, fix, owner; exit 3 on an alert
scripts/fleet findings /tmp/PROJ-12-review.md          # only the findings of a review transcript, deduplicated
scripts/fleet watch --stop-prefix PROJ-12-  # stop every watcher of one ticket, before the report
scripts/fleet free-models refresh          # today's free models, for read-only scans
scripts/fleet stats --role engineer        # the executor scoreboard

The token counts the log keeps for a sub-session are the size of its last call, not its spend; cost reads the spend from the session files. Engineering's own tools (board drift, pre-pick triage, git tidy, the leftover sweep) are in its scripts folder.

Project config

Each repo carries .plane/config.json (Plane project, trunk, quality gate, CI, deploy, slots). Private values (client contact, server names, timezone) go in a per-user override outside the repo. When a repo has no config yet, the PM and solo roles build it from the board and the repo.

Updates

This skill is versioned. Its version is metadata.version in the header of this file. CHANGELOG.md in this folder lists what changed in each version, newest first.

To check for a newer version, open the address on the Latest: line of CHANGELOG.md and compare its top version with the installed one. If the newer one is ahead, read every entry between the two and tell the user what changed before anything is updated. A Breaking section means the user has to do something. To update, reinstall the skill from its source repository, the same way it was installed.

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.