test-fixing
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (LOW): The skill executes shell commands including 'make test' and 'uv run pytest' to identify and verify failures. While these are necessary for fixing tests, they execute code within the project environment. Severity is downgraded as this is the primary purpose of the skill.
- [PROMPT_INJECTION] (LOW): Vulnerable to Indirect Prompt Injection. The agent is instructed to 'Analyze output' from test runs. If a test case or the code under test outputs malicious instructions, the agent may follow them. * Evidence: (1) Ingestion points: Output from 'make test' and 'pytest'. (2) Boundary markers: Absent. (3) Capability inventory: Command execution via shell and file modification via Edit tool. (4) Sanitization: Absent.
- [DATA_EXPOSURE] (SAFE): Reads local project files and 'git diff' to diagnose errors. No evidence of external exfiltration or access to sensitive global paths like ~/.ssh.
Audit Metadata