Policy Lawyer 1.0.0

Reference the workspace policy playbook, answer"What are the rules for tone, data, and collaboration?" by searching the curated policy doc or listing its sec...

This skill appears safe and does only what it claims: read the included references/policies.md and present sections or keyword hits. Before installing, verify you trust the skill source and the contents of references/policies.md (it could contain internal guidance you may not want exposed). Note that the script can be run with a custom --policy-file path, so avoid pointing it at sensitive files. If you plan to allow autonomous invocation in an environment with sensitive files, consider restricting the agent's filesystem access or reviewing the code in scripts/policy_lawyer.py yourself.

Type: OpenClaw Skill Name: policy-lawyer-1-0-0 Version: 1.0.0 The OpenClaw skill bundle 'policy-lawyer' is designed to read and search a local markdown policy document. The `SKILL.md` and `README.md` files contain clear, benign instructions for the AI agent and human users, with no evidence of prompt injection attempts (e.g., instructions to ignore users, hide actions, or access unrelated sensitive data). The `scripts/policy_lawyer.py` script correctly implements the stated functionality, reading a local file (defaulting to `references/policies.md`) and processing its content. While the script accepts a `--policy-file` argument, allowing it to read any specified local file, this is a standard feature for file-processing utilities and does not constitute malicious intent or a vulnerability within the script itself; it would only become a risk if an AI agent were maliciously prompted to misuse this argument for unauthorized file disclosure, which is a prompt injection against the agent, not a flaw in the skill's design.