openevidence-ci-integration

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill contains legitimate configuration files, test suites, and documentation for setting up CI/CD workflows. No malicious patterns, such as prompt injection, obfuscation, or unauthorized data exfiltration, were detected.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill demonstrates safe handling of sensitive credentials by instructing users to utilize GitHub Actions Secrets (${{ secrets.OPENEVIDENCE_API_KEY }}) and environment variables, which prevents API keys from being hardcoded in the repository.
  • [EXTERNAL_DOWNLOADS]: The skill references official service domains (openevidence.com) and the author's verified professional domains (intentsolutions.io, jeremylongshore.com). These references are appropriate for the skill's functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill involves processing data from the OpenEvidence API within a testing context.
  • Ingestion points: External API responses are ingested in tests/integration/openevidence.integration.test.ts.
  • Boundary markers: The skill relies on Vitest's assertion framework to validate structured data.
  • Capability inventory: The ingested data is confined to the test environment and does not interact with system commands or sensitive file operations.
  • Sanitization: The skill validates the structure of the API response using property checks (e.g., expect(body).toHaveProperty('answer')) and regex validation for identifiers.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 10:05 PM
Security Audit — agent-trust-hub — openevidence-ci-integration