Osop Review

Review .osop/.osoplog for security risks, permission gaps, and destructive commands

This skill appears to do what it says (analyze .osop/.osoplog contents) but asks for access to ~/.osop/config.yaml and declares bash as required without justifying either. Before installing or running it: - Ask the skill author why ~/.osop/config.yaml is required and what data from it will be read. If the skill needs config context, it should state that in SKILL.md and explain exactly what keys are used. - Confirm whether the skill will actually read the config path or only the argument file. If uncertain, run the skill on a copy of your .osop/.osoplog in a sandbox or with a wrapper that prevents reading ~/.osop. - Verify the bash requirement: if the skill is purely a file analyzer, it shouldn't need a shell binary. Prefer a skill that lists only the minimal dependencies. - When running, avoid supplying real secrets or production logs; test first with redacted samples. If the author updates SKILL.md to explicitly explain the use of ~/.osop/config.yaml (or removes that requirement) and documents any shell commands it runs, the assessment would likely move to benign. If the skill actually reads the config for tokens or other secrets without clearly declaring that, treat it as high-risk and do not install.

Type: OpenClaw Skill Name: osop-review Version: 1.2.0 The skill bundle contains instructions for an AI agent to perform security audits on OSOP workflow (.osop) and log (.osoplog.yaml) files. It specifically directs the agent to identify high-risk commands (e.g., 'rm -rf'), hardcoded secrets (e.g., GitHub tokens), and missing security controls like approval gates or timeouts. There is no executable code or evidence of malicious intent; the instructions are purely defensive and aligned with the stated purpose of security reviewing.