api-test-suite-builder
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of OpenAPI/Swagger JSON specifications to generate executable test code. This creates an attack surface where a malicious specification could attempt to influence the agent's behavior or the generated output.
- Ingestion points: The
scripts/test_generator.pyscript reads and parses JSON files provided by the user or found in the workspace. - Boundary markers: The skill does not use specific delimiters or warnings to isolate content from the OpenAPI spec when generating code docstrings and identifiers.
- Capability inventory: The generator script has the capability to write files to the local filesystem via the
--outputargument. - Sanitization: While JSON payloads are properly serialized, metadata from the spec (such as
summaryoroperationId) is interpolated into Python/TypeScript code templates. Specifically,test_generator.pyinserts thesummaryfield directly into triple-quoted docstrings without escaping, which could allow for code injection if a malicious spec contains triple-quotes followed by arbitrary code. - [DYNAMIC_EXECUTION]: The skill generates Python and TypeScript test scripts at runtime based on templates and provided specifications. While this is the intended functionality of a test generator, the resulting code's logic is dynamically derived from external input.
Audit Metadata