test
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes project test suites using shell commands such as
go test ./...andnpm testbased on identified file changes. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data that could contain malicious instructions. 1. Ingestion points: Git commit history and test output (SKILL.md). 2. Boundary markers: Absent; the skill does not instruct the agent to ignore instructions embedded in logs or commit messages. 3. Capability inventory: Execution of shell commands for testing (SKILL.md). 4. Sanitization: Absent.
Audit Metadata