Review Ai Writing

This skill appears to do what it says (scan repo text for AI-style writing) but has a few inconsistencies you should consider before installing or running it: - Declared vs. actual runtime tools: SKILL.md expects to run git and the GitHub CLI ('gh') and to spawn subagents via a Task tool, but the skill metadata lists no required binaries. Ensure the environment you run this in has git/gh available and that you accept those tools being invoked. - GitHub authentication: If you run this where the gh CLI is authenticated, the skill will be able to read PR descriptions and other data accessible to that account. If you don't want it to access PR bodies, run it in an environment without gh auth or avoid running the PR-scanning mode. - File I/O: The skill will read repository files (including .beagle/llm-artifacts-review.json if present) and will write its report to .beagle/ai-writing-review.json. Review that output before committing or sharing it. - Subagents and other skills: The workflow loads other skills (e.g., beagle-core:review-verification-protocol and possibly itself under a different namespace) and spawns subagents. Confirm you trust those other skills and your agent platform's Task tool, since they will process repository contents as well. If you plan to use this skill, ask the provider to: (1) list required binaries (git, gh, find, etc.) in the metadata, (2) document exactly what other skills it loads and why, and (3) clarify any network or credential usage. If you cannot verify those, run it in an isolated environment or without gh authentication.