project-test-loop
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- COMMAND_EXECUTION (LOW): The skill dynamically detects and executes test commands (e.g.,
npm test,pytest,cargo test) based on project files likepackage.jsonorMakefile. In an adversarial context, a malicious repository could define dangerous commands in these files that would be executed by the agent via theBashtool. - PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8). It ingests and parses the output of test commands to identify fixes and refactoring opportunities.
- Ingestion points: Failure output and assertion results from
Bashcommand execution. - Boundary markers: Absent. The skill does not specify delimiters or instructions to ignore embedded commands in the test output.
- Capability inventory:
Read,Edit, andBashtools allow for file system modification and further command execution. - Sanitization: Absent. There is no logic mentioned for escaping or validating the content of the test output before it is processed by the LLM logic.
Audit Metadata