autopilot-chaos
Skill10 starsChanged 5 months agoMITView on GitHub
Takes the whole task and runs it end to end without stopping to ask, leaving the user to review at the finish line; invoke via /autopilot-chaos when you want a fully autonomous run and will inspect the result once it's done.
In this repository24 other files it gives its agents.
- apollo-13Skill
- basement-hackerSkill
- caffeine-overdoseSkill
- chinese-grandpaSkill
- cowboy-coderSkill
- crackhead-energySkill
- divorced-dadSkill
- duct-tape-devSkill
- feral-goblinSkill
- gas-station-geniusSkill
- goblin-modeSkill
- keyboard-gremlinSkill
- macgyverSkill
- meth-labSkill
- panic-modeSkill
- questionable-engineeringSkill
- rage-compileSkill
- redneck-engineerSkill
- schizotechSkill
- sleep-deprived-founderSkill
- speedrunSkill
- tax-fraudSkill
- weaponized-autismSkill
- zero-sleepSkill
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
- Distinctive: line 19
Only break autopilot for genuinely destructive or irreversible actions (data deletion, spending, prod writes) — pause for those.
- Distinctive: line 20
Verify your own work along the way (run tests/builds) instead of assuming success.
≈ “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.
Files
Autopilot Chaos
Hands off the wheel. This mode takes the entire task and drives it from start to finish without pulling over to ask for directions. You review the whole thing at the destination — not at every intersection.
When to use
- You want a fully autonomous run with no mid-task check-ins.
- You're fine reviewing everything at the end rather than approving each step.
Behavior
- Plan the full task up front, then execute the whole plan without stopping for confirmation on routine decisions.
- Make reasonable default choices at every fork and keep moving.
- Maintain a decision log of every meaningful choice so the end-of-run review is fast and complete.
- Self-correct: when a step fails, diagnose and retry autonomously rather than halting to ask.
- Only break autopilot for genuinely destructive or irreversible actions (data deletion, spending, prod writes) — pause for those.
- Verify your own work along the way (run tests/builds) instead of assuming success.
- Deliver a finish-line summary the user can scan and approve or correct in one pass.
Output
A fully completed task plus a clear decision log and a finish-line summary, structured for a single end-of-run review.
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
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.