simplify-and-harden

Post-completion self-review for coding agents that runs simplify, harden, and micro-documentation passes on non-trivial code changes. Use when: a coding task...

This skill appears coherent and low-risk, but take these precautions before enabling it: 1) If you run the documented 'npx skills add' command, inspect the remote package (pskoett/pskoett-ai-skills) first — npx will fetch and run external code. 2) Be aware interactive mode can apply 'simple patches' automatically; if you want to review every change, disable automatic edits or run the CI-only variant. 3) Confirm your agent actually enforces the 'only modified files' and budget/time limits to avoid unintended edits to unrelated code. 4) Keep independent reviews in your workflow — the skill is explicitly a first-pass self-review, not a substitute for external review. 5) Try the skill on small, low-risk tasks first and verify generated diffs/commits before using it on critical repositories.

Type: OpenClaw Skill Name: simplify-and-harden Version: 1.0.1 The OpenClaw AgentSkills skill bundle 'simplify-and-harden' is designed for post-completion code quality and security review. All files, including `SKILL.md` and `references/agent-context-snippets.md`, consistently instruct the AI agent to operate within strict scope and budget constraints, only modifying code it previously touched. Crucially, any structural changes (refactors or security refactors) require explicit human approval, with the agent instructed to pause and wait. The self-improvement loop is designed to refine the agent's own internal prompts, not to exfiltrate data or establish persistence. There is no evidence of malicious intent, data exfiltration, unauthorized execution, or prompt injection designed to bypass safety measures; instead, the instructions reinforce safety and human oversight.