How do I install the Fleet Execute skill?

Run npx skillmds@latest add damon-stewart-1/fleet-execute in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Fleet Execute safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Fleet Execute?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Fleet Execute free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Fleet Execute?

Damon-Stewart-1 (@damon-stewart-1) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Fleet Execute by damon-stewart-1 — Agent Skill

Name: Fleet Execute
Author: Damon-Stewart-1

Fleet Execute

Execute READY plans from the fleet triage queue. Shows which plans scored ready, asks for confirmation, then dispatches agents. Triggers on: /fleet-execute, 'run fleet', 'execute ready plans', 'dispatch fleet', 'run the fleet'.

File contents

/fleet-execute: Plan Fleet Dispatcher

Dispatches autonomous agents to execute plans that scored READY in the last triage run.

Setup

This skill assumes:

A triage output directory, default ~/Claude-Stuff/fleet/ (adjust to your setup)
A fleet-execute.sh script at ~/bin/fleet-execute.sh
A fleet-triage.sh script at ~/bin/fleet-triage.sh

Both scripts must be executable. Adjust paths in this skill to match your local install.

What to do

Check triage data. Read ~/Claude-Stuff/fleet/triage-latest.json and ~/Claude-Stuff/fleet/last-triage.json. If no file exists or triage is >12 hours old, tell the user and suggest running ~/bin/fleet-triage.sh first.
Show READY plans. Display each plan that scored READY (score 10-12), with:
- Plan name
- Score (X/12)
- Per-criterion breakdown (acceptance / no-human / paths / idempotency)
- One-sentence summary
- Any noted blockers
Confirm with user. Ask: "Ready to dispatch agents for these N plans? (yes/no, or name specific plans to skip)"
On confirmation: Run:
```
~/bin/fleet-execute.sh
```
Do NOT pass --autonomous. Interactive mode only here.
Tell the user:
- Run output lands at: ~/Claude-Stuff/fleet/runs/{date}-{time}/
- Escalations go to: ~/Claude-Stuff/fleet/needs-human.md
After completion: Read the latest ~/Claude-Stuff/fleet/runs/*/summary.md and report: how many completed, how many escalated.

Stale triage handling

If last-triage.json is missing or triage was >12 hours ago:

"No recent triage data (last run: {date} or never). Run ~/bin/fleet-triage.sh manually. Triage takes 2-5 minutes."

Hold file

To cancel today's execution without running this skill:

touch ~/Claude-Stuff/fleet/hold
# or for a specific date:
touch ~/Claude-Stuff/fleet/hold-$(date +%Y-%m-%d)

Dry run

To see what would be dispatched without running agents:

~/bin/fleet-execute.sh --dry-run

Notes

Agents run with --allowedTools "Read,Glob,Grep,Edit,Write" (no Bash in allowedTools)
Each agent has a 30-minute timeout
Max 2 agents run concurrently
Failed or blocked agents escalate to needs-human.md automatically

---name: fleet-executedescription: Execute READY plans from the fleet triage queue. Shows which plans scored ready, asks for confirmation, then dispatches agents. Triggers on: /fleet-execute, 'run fleet', 'execute ready plans', 'dispatch fleet', 'run the fleet'.---# /fleet-execute: Plan Fleet DispatcherDispatches autonomous agents to execute plans that scored READY in the last triage run.## SetupThis skill assumes:- A triage output directory, default `~/Claude-Stuff/fleet/` (adjust to your setup)- A `fleet-execute.sh` script at `~/bin/fleet-execute.sh`- A `fleet-triage.sh` script at `~/bin/fleet-triage.sh`Both scripts must be executable. Adjust paths in this skill to match your local install.## What to do1. **Check triage data.** Read `~/Claude-Stuff/fleet/triage-latest.json` and `~/Claude-Stuff/fleet/last-triage.json`. If no file exists or triage is >12 hours old, tell the user and suggest running `~/bin/fleet-triage.sh` first.2. **Show READY plans.** Display each plan that scored READY (score 10-12), with: - Plan name - Score (X/12) - Per-criterion breakdown (acceptance / no-human / paths / idempotency) - One-sentence summary - Any noted blockers3. **Confirm with user.** Ask: "Ready to dispatch agents for these N plans? (yes/no, or name specific plans to skip)"4. **On confirmation:** Run: ```bash ~/bin/fleet-execute.sh ``` Do NOT pass `--autonomous`. Interactive mode only here.5. **Tell the user:** - Run output lands at: `~/Claude-Stuff/fleet/runs/{date}-{time}/` - Escalations go to: `~/Claude-Stuff/fleet/needs-human.md`6. **After completion:** Read the latest `~/Claude-Stuff/fleet/runs/*/summary.md` and report: how many completed, how many escalated.## Stale triage handlingIf `last-triage.json` is missing or triage was >12 hours ago:> "No recent triage data (last run: {date} or never). Run `~/bin/fleet-triage.sh` manually. Triage takes 2-5 minutes."## Hold fileTo cancel today's execution without running this skill:```bashtouch ~/Claude-Stuff/fleet/hold# or for a specific date:touch ~/Claude-Stuff/fleet/hold-$(date +%Y-%m-%d)```## Dry runTo see what would be dispatched without running agents:```bash~/bin/fleet-execute.sh --dry-run```## Notes- Agents run with `--allowedTools "Read,Glob,Grep,Edit,Write"` (no Bash in allowedTools)- Each agent has a 30-minute timeout- Max 2 agents run concurrently- Failed or blocked agents escalate to `needs-human.md` automatically

Frequently asked questions