issue
Pass
Audited by Gen Agent Trust Hub on Apr 25, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes the GitHub CLI (gh) for issue and label creation. These actions are legitimate for the skill's intended purpose and are restricted to specific subcommands by the platform's tool configuration, adhering to the principle of least privilege.
- [PROMPT_INJECTION]: The skill processes project data and user input to generate content, creating a potential surface for indirect prompt injection. This is a standard characteristic for this type of tool and is mitigated by the structural use of templates and heredoc string handling. Ingestion points: local repository files and user conversation context. Boundary markers: markdown issue templates. Capability inventory: repository resource management via the gh CLI. Sanitization: employment of the cat <<'EOF' pattern to encapsulate multi-line issue content, preventing shell expansion of ingested data.
Audit Metadata