api-testing

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill references several command-line tools for API development and testing, such as prism for mocking, k6 for load testing, oasdiff for schema comparison, and newman for Postman collection execution. These are standard tools for the described workflow and are used safely within the provided instructions.
  • [EXTERNAL_DOWNLOADS]: The skill incorporates well-known libraries and frameworks from official registries, including @pact-foundation/pact, msw, @pollyjs/core, and ajv from NPM, as well as locust from PyPI. These dependencies are standard in the testing ecosystem.
  • [DATA_EXFILTRATION]: Network requests (e.g., via fetch, k6/http, or locust) are used exclusively for their intended purpose of testing local or specified API endpoints. No unauthorized data transmission or exfiltration patterns were detected.
  • [CREDENTIALS_UNSAFE]: The skill uses placeholder values (e.g., abc123, {{authToken}}) for tokens and example URLs (e.g., pact-broker.company.com). It correctly promotes the use of environment variables for managing sensitive configuration in testing environments.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 09:13 PM
Security Audit — agent-trust-hub — api-testing