The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The installation script (install.sh) and instructions in the README.md employ standard patterns to download and execute code from the author's official GitHub repository and the public NPM registry. These are trusted sources for the tool's distribution.\n- [COMMAND_EXECUTION]: The skill documentation in SKILL.md utilizes the !command syntax to retrieve help data for the glab CLI. These commands are used solely for information retrieval and are not part of any malicious execution chain.\n- [SAFE]: The project includes a comprehensive set of security enforcement scripts located in hooks/claude/. These scripts, such as block-sensitive-files.sh and sanitize-output.sh, provide significant defense-in-depth by monitoring the agent's actions to prevent unauthorized access to sensitive local files and accidental exposure of credentials.

mayar