gh-agent-task
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the
github/agent-taskextension through the official GitHub CLI. This is a trusted vendor resource. - [COMMAND_EXECUTION]: The extension utilizes the
ghcommand-line interface to interact with GitHub issues and pull requests as part of its documented functionality. - [PROMPT_INJECTION]: The skill defines an attack surface for indirect prompt injection by processing natural language instructions from GitHub issues. 1. Ingestion points: GitHub issue descriptions and metadata via the
gh-awworkflow integration. 2. Boundary markers: Utilizes a 'staged' mode and 'safe-outputs' configuration to preview and control agent behavior. 3. Capability inventory: Ability to create issues, update descriptions, and trigger automated pull requests via the GitHub CLI. 4. Sanitization: Relies on the user's manual review of generated code and the underlying Copilot service's safety filters.
Audit Metadata