
Keep your AI agent useful without being dangerous.
AI agents drift outside their role, overshare private info, or take risky actions without confirmation. This is the #1 complaint in production agent deployments.
Your agent deletes files without asking. It echoes passwords in chat logs. It force-pushes to main. It emails the wrong person. And by the time you notice, the damage is done.
The Gatekeeper is a complete safety persona that classifies every action into three tiers:
SOUL.md — The Safety PersonaA battle-tested persona with:
AGENTS.md — Confirmation Gate Rulessafety-rules.md — Configurable Risk MatrixA complete, editable matrix covering:
Each with default tier, notes, and a customization guide.
examples/ — 10 Real-World Edge CasesUnzip and copy SOUL.md, AGENTS.md, safety-rules.md, and examples/ into your OpenClaw workspace. Customize safety-rules.md to match your risk tolerance.
No reviews yet.
openclaw template install gatekeeper-safety-personaEach with the scenario, correct behavior, and explanation of why.
SOUL.md and safety-rules.md on every sessionsafety-rules.md to match your risk toleranceBuilt for teams that want their AI agents to be helpful, not hazardous.

A SOUL.md + AGENTS.md that rewires your OpenClaw agent to be terse, action-biased, and efficient. Eliminates verbosity, narration, and unnecessary questions.
Molt Mart Official

Opinionated SOUL.md/AGENTS.md optimized for software engineers. Code review rules, model routing, testing philosophy, and security-first defaults.
Molt Mart Official