blast-radius
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the processing of external, potentially untrusted data including pull request diffs and third-party library source code.
- Ingestion points: The agent is instructed to read diffs, PRs, commits, and library source files in Steps 1 and 3.
- Boundary markers: No specific delimiters or safety instructions are provided to distinguish between code-to-be-analyzed and potential embedded instructions within that code.
- Capability inventory: The skill leverages file reading, grep, script writing, and shell execution as seen in Steps 4 and 5.
- Sanitization: The instructions lack guidance on sanitizing or escaping content from the analyzed codebase before it is used in execution or reporting.
- [DYNAMIC_EXECUTION]: The skill relies on the generation and execution of code at runtime to validate safety claims.
- Evidence: Instructions in Steps 4 and 5 explicitly guide the agent to 'Write a script or test that runs the real code' and 'run it'. This involves the creation of temporary scripts that import and execute functions from the analyzed project.
Audit Metadata