test-fixing
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, such as test failure logs, CI configurations, and dependency manifests, which serves as a potential vector for indirect prompt injection.
- Ingestion points: Project instructions, dependency manifests, CI configuration, and test output analyzed in
SKILL.md. - Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' markers when reading external project files or logs.
- Capability inventory: The skill utilizes file editing tools and shell command execution (
make test,uv run pytest) as part of its fixing and verification process. - Sanitization: There is no evidence of sanitization, filtering, or validation of the content ingested from the external project files or test outputs.
Audit Metadata