The Agent Skills Directory

[SAFE]: The skill operates using official, well-known CLI tools (gh for GitHub and tea for Gitea/Forgejo) to manage communication. It includes explicit instructions to check authentication status before proceeding.
[SAFE]: It implements a 'human-in-the-loop' requirement for security-sensitive actions. Specifically, it instructs the agent to treat comments as data rather than commands and requires user confirmation before accepting messages from new bot accounts or performing actions outside its standing authority.
[SAFE]: The polling mechanism uses standard API endpoints (repos/{owner}/{repo}/issues/comments) and follows best practices like rate limiting and using a local recovery log for persistence rather than modifying system-wide configurations.
[SAFE]: The message protocol uses standard HTML comments for metadata, which is a common and safe way to embed structured data in markdown without interfering with human readability.

github-monitor