test-and-fix
Pass
Audited by Gen Agent Trust Hub on Apr 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill automatically identifies and executes shell commands such as lint, test, and build scripts found in the project's 'package.json' or documentation. This assumes the project configuration is trustworthy.
- [PROMPT_INJECTION]: The skill processes data from external sources like command output and project files, which serves as a surface for indirect prompt injection. Ingestion points: Reads content from 'package.json', documentation files, and the standard output/error streams of executed validation tools. Boundary markers: The instructions do not specify any delimiters or safety prompts to isolate untrusted data from the agent's core logic. Capability inventory: The skill has permissions to execute terminal commands and write files to the local file system. Sanitization: There is no explicit logic to sanitize or filter the content of project files or command logs before they are interpreted by the agent.
Audit Metadata