generating-unit-tests
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (HIGH): Significant vulnerability surface exists because the skill ingests untrusted data (source code and git diffs) and possesses high-privilege capabilities (writing files and executing shell commands). ● Ingestion points: Reading file content and git diffs in Step 1 and Step 3. ● Boundary markers: No delimiters or ignore instructions are used. ● Capability inventory: Executes npx, npm, and git commands; writes new test files. ● Sanitization: No sanitization performed.
- [Command Execution] (HIGH): The skill runs shell commands such as npx jest and npm install. Maliciously crafted file names or repository content could be used for command injection.
- [External Downloads] (LOW): Suggests dynamic installation of packages. While standard, this is a risk if package names are manipulated.
Recommendations
- AI detected serious security threats
Audit Metadata