setup-matt-pocock-skills
Pass
Audited by Gen Agent Trust Hub on Jun 24, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Uses
git remoteto identify the repository host and provides instructions for the agent to useghorglabCLI tools for issue management activities. - [PROMPT_INJECTION]: The skill ingests data from local repository files and remote issue tracker comments, creating an indirect injection surface.
- Ingestion points: Reading
AGENTS.md,CLAUDE.md, and issue data viagh/glabinSKILL.md. - Boundary markers: Absent.
- Capability inventory: File system writes to
docs/agents/and command execution through supported issue tracker CLIs. - Sanitization: Mitigated by mandatory human-in-the-loop verification where the user must approve all proposed content before files are written.
Audit Metadata