- name
- agent-bootstrap-hardening
- description
- Audit and harden an OpenClaw workspace setup (AGENTS/SOUL/USER/HEARTBEAT/memory) with concise, enforceable rules, safety boundaries, and low-bloat instruction design. Use when users ask to optimize, clean up, or harden their agent brain files.
Agent Bootstrap Hardening
Use this skill when a user wants to improve core workspace files for better reliability, speed, and safety.
Scope
Target files (if present):
AGENTS.mdSOUL.mdUSER.mdHEARTBEAT.mdMEMORY.mdmemory/YYYY-MM-DD.md- optional helper docs:
RESOLVER.md,QUALITY.md
Workflow
- Inventory current files and obvious gaps.
- Score each file quickly with the rubric in
references/rubric.md. - Harden content with minimal edits:
- keep instructions concise and operational - tighten safety boundaries and approval rules - reduce duplicated or theory-heavy text - preserve user tone/persona intent
- Add missing primitives when useful:
- RESOLVER.md for execution routing - QUALITY.md for lightweight pre-send checks
- Return summary:
- what changed - why it improved behavior - remaining risks/todos
Editing rules
- Prefer small, targeted diffs over wholesale rewrites.
- Avoid introducing process that cannot be followed consistently.
- Keep AGENTS and SOUL readable in one pass.
- Preserve explicit user preferences and platform-specific norms.
Guardrails
- Never weaken privacy or approval boundaries.
- Never add hidden/external exfil behavior.
- Do not claim setup is “secure” without caveats.
- If uncertain about impact, ask before applying broad policy changes.
Output contract
Provide:
- Hardening summary (3–8 bullets)
- Files changed list
- Next recommended step (single highest-leverage follow-up)