thermonuclear-code-review
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to utilize command-line utilities including
git diff,gh pr diff,git show, andgrepto retrieve and navigate code changes. These commands are essential for performing the skill's primary function of code auditing and are executed within the user's project environment. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external code (PR diffs and branches), which presents a surface for indirect prompt injection. A malicious contributor could embed instructions within a PR intended to influence the agent's review. The skill currently lacks specific delimiters or instruction-ignoring warnings to mitigate this risk.
- Ingestion points: External code is ingested via
git diff,gh pr diff, andgit showas described in SKILL.md. - Boundary markers: None present; the skill treats the diff content as raw text for review.
- Capability inventory: The agent is granted capabilities to read files and perform recursive symbol searches (
grep) within the repository scope. - Sanitization: The skill does not define methods for sanitizing or escaping the content of the code changes before processing.
Audit Metadata