property-based-testing
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill identifies and mitigates a potential attack surface where untrusted data from test outputs enters the agent context.\n
- Ingestion points: The agent reads logs and output emitted by software dependencies during test execution, which are noted as untrusted in
references/frameworks/jqwik.md.\n - Boundary markers: Explicit instructions are provided to the agent to treat external prose as untrusted and maintain user/repository instructions as the authoritative source of truth.\n
- Capability inventory: The skill utilizes standard project tooling such as
pytest,npm test,cargo test, andgo testfor operating properties.\n - Sanitization: The guidelines emphasize environmental isolation, deterministic cleanup, and isolation of mutable generated values to manage state and prevent side effects (
references/stateful-and-concurrent-testing.md).\n- [SAFE]: The skill references official documentation for established testing frameworks including Hypothesis, fast-check, jqwik, and proptest. These external links target trusted domains (e.g., hypothesis.readthedocs.io, fast-check.dev, jqwik.net) for legitimate technical guidance.
Audit Metadata