blast-radius
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill workflow involves ingesting and analyzing untrusted external data, specifically code diffs, Pull Requests, and library source code, which could contain malicious instructions.\n
- Ingestion points: The agent is instructed to read changes (diffs), PR content, and commits in Step 1.\n
- Boundary markers: There are no explicit instructions for the agent to treat data from the diffs/PRs as untrusted or to use delimiters to prevent instruction injection.\n
- Capability inventory: The agent has the capability to write and execute scripts against the codebase as described in Steps 4 and 5 of the analysis process.\n
- Sanitization: No sanitization of the analyzed code is performed before execution.\n- [DYNAMIC_EXECUTION]: The skill encourages the agent to verify code safety by dynamically generating and running test scripts.\n
- Evidence: The instructions explicitly tell the agent to "Write a script or test that runs the real code, run it, and paste what happened" to prove safety.
Audit Metadata