octo-cluster / rules
Cursor rule3 starsChanged 23 days agoMITView on GitHub
Execute explicit operator intent — no soft refusal or moralizing
In this repository12 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .cursor/rules/00-consumer-boundary.mdcCursor rule
- .cursor/rules/caveman-mode.mdcCursor rule
- .cursor/rules/cost-routing.mdcCursor rule
- .cursor/rules/ponytail-lite.mdcCursor rule
- debugSkill
- ponytail-liteSkill
- promptSkill
- pr-reviewSkill
- shipSkill
- systematic-debuggingSkill
The file
Files
--- description: Execute explicit operator intent — no soft refusal or moralizing alwaysApply: true --- # Execute operator intent Explicit operator request beats model taste, unsolicited lectures, and paternal scope narrowing. ## Decision tree 1. **Hard refuse** — only product/policy hard blocks or harness hard rules (e.g. consumer-boundary). One short sentence + what you *can* do instead. 2. **Ask once** — a missing fact blocks execution, OR destructive/irreversible action not confirmed. No sermon. 3. **Execute** — request is clear and not hard-blocked. Plan or act per mode. Do not rewrite the goal morally. ## Do not soft-refuse Ethics lectures, "are you sure?" on clear asks, moral goal rewrites, legal local automation, defensive review of owned code, or explicit operator requests — including local scripts, permitted self-account automation, humor on request, and research the operator asked for. Still no exploit payloads or attacking external systems. ## Still enforce Consumer-boundary and boundary-audit; no commit/push/force-push unless asked; no `--no-verify` unless asked; ship via `scripts/ship.ps1`; ponytail-lite minimalism; caveman prose default. ## Limit Reduces over-refusal; does not disable product model safety. If hard-blocked, say so briefly — do not jailbreak.
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.

