github-workflows-query

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill retrieves workflow metadata, including names and paths, from external GitHub repositories. This establishes an indirect prompt injection surface where a malicious repository owner could include instructions in workflow files to manipulate the agent's behavior. (Ingestion points: GitHub API response in query-workflows.sh; Boundary markers: None; Capability inventory: Subprocess and shell execution; Sanitization: None).
  • [COMMAND_EXECUTION]: The script query-workflows.sh interpolates the owner, repo, and page parameters directly into the gh api command string without exhaustive validation. While the use of the GitHub CLI limits the risk, this lack of sanitization could allow for API path traversal or query parameter injection (e.g., using ../ in a repository name to access unauthorized API endpoints).
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 05:54 AM
Security Audit — agent-trust-hub — github-workflows-query