gh-project-board

Pass

Audited by Gen Agent Trust Hub on Jun 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the GitHub CLI (gh) to query and update GitHub Project boards. It executes the CLI via execFileSync using argument arrays, which prevents shell injection risks.
  • [DATA_EXFILTRATION]: No unauthorized data transmission detected. The skill reads project metadata and writes configuration changes to the GitHub API via the local gh utility as part of its documented workflow.
  • [PROMPT_INJECTION]: The skill features an indirect prompt injection surface (Category 8) by processing data from external GitHub projects:
  • Ingestion points: Project titles, field names, and option descriptions are retrieved from GitHub in scripts/setup-gh-project-board.mjs.
  • Boundary markers: Not applicable as data is processed by script logic.
  • Capability inventory: The script can create and update single-select fields on GitHub projects.
  • Sanitization: The script uses JSON.stringify to escape string values before they are used in GraphQL mutations, mitigating injection into the API requests.
  • [SAFE]: No obfuscation techniques, hardcoded credentials, or persistence mechanisms were found. The skill's behavior is transparent and aligns with its described functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 20, 2026, 01:13 PM
Security Audit — agent-trust-hub — gh-project-board