developer-community-health
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process data from various external sources, such as Git history, Forge APIs, and community venue exports, to generate health reports. It lacks explicit instructions for the agent to sanitize this data or use boundary markers when interpolating it into prompts or reports. This creates a surface for indirect prompt injection where malicious content in a repository or API response could attempt to manipulate the agent's analysis or subsequent actions.\n
- Ingestion points: Local and remote Git repositories, Forge APIs (GitHub/GitLab), and venue-specific analytics exports.\n
- Boundary markers: No specific delimiters or "ignore" instructions are provided to scope external content.\n
- Capability inventory: Access to local file systems, Git commands, and execution of shell scripts.\n
- Sanitization: No explicit sanitization or validation steps for external data are included in the instructions.\n- [COMMAND_EXECUTION]: The skill includes a bash utility (
scripts/contributor-absence-factor.sh) and directs the agent to execute it for metric calculation. The script takes user-influenced arguments, such as regex filters for excluding contributors, which are passed to shell commands. A lack of strict input validation within the script or the agent's instructions could potentially lead to command injection if malicious strings are passed as arguments to the script's shell environment.
Audit Metadata