litestar-testing
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides examples for testing API handlers that ingest untrusted data from web requests (e.g., create_user in references/failure-patterns.md). This constitutes an attack surface if the agent processes malicious payloads through these tests. Ingestion points: The create_user and search handlers in references/failure-patterns.md receive data via Request objects or query parameters. Boundary markers: None defined in the reference snippets. Capability inventory: The skill utilizes subprocess calls for live-server testing in references/client-patterns.md. Sanitization: The examples focus on test structure rather than input validation logic.
- [COMMAND_EXECUTION]: The documentation in SKILL.md and references/client-patterns.md encourages the use of subprocess_sync_client and subprocess_async_client. These Litestar utilities spawn external server processes to facilitate integration testing for specific transport protocols like SSE.
Audit Metadata