blast-radius
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill explicitly instructs the agent to verify code safety by writing and running scripts or tests that call the application's real code (Steps 4 and 5). This capability involves the runtime generation and execution of code based on the agent's interpretation of the project.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, specifically pull request diffs, commits, and symbols (Step 1). Since this data influences the generation of verification scripts (Step 5), there is a risk that a malicious diff could be crafted to exploit the agent's logic or induce unsafe execution.
- Ingestion points: Pull request bodies, diff content, and commit messages (SKILL.md Step 1).
- Boundary markers: Absent; the skill does not define specific delimiters for untrusted content.
- Capability inventory: File writing and command execution via shell scripts or test runners.
- Sanitization: Absent; while the skill advises stripping private data from output, it lacks sanitization for the inputs used during script generation.
Audit Metadata