prompt-regression-tester
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a framework for processing external data through test cases, which creates a surface for indirect prompt injection.
- Ingestion points: Untrusted data enters the agent context through the 'input' and 'context' fields in the test_cases JSON structure defined in SKILL.md.
- Boundary markers: The implementation uses standard string interpolation for prompt construction without explicit boundary markers or instructions to ignore embedded commands.
- Capability inventory: The logic is restricted to LLM invocation and report generation; no shell execution, file-write operations, or network requests are present.
- Sanitization: No sanitization or validation is applied to the test case inputs before interpolation.
Audit Metadata