post-dev-test
Warn
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
Bashtool to run verification tests. This allows for arbitrary command execution within the environment context usingyarn jest.\n - Evidence: Phase 5 and the 'Execute Tests' section in
SKILL.mdshow the workflow of using theBashtool to run test suites.\n- [REMOTE_CODE_EXECUTION]: The skill follows a pattern of dynamic code generation and execution by writing test files to the filesystem and then running them. This represents a dynamic execution risk factor.\n - Evidence:
SKILL.mdPhase 4 ('Write Tests') and Phase 5 ('Execute Verification') document the process of writing new executable scripts and then invoking them.\n- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it reads untrusted project data to inform its script generation logic.\n - Ingestion points: Phase 2 in
SKILL.mdreads project files viaRead,Grep, andGlobtools to assess coverage gaps.\n - Boundary markers: There are no explicit boundary markers or 'ignore' instructions for the data ingested from the source code.\n
- Capability inventory: The agent has high-privilege capabilities including
WriteandBashwhich could be misused if influenced by malicious content in analyzed files.\n - Sanitization: No sanitization is applied to the code content extracted from the project before it is used to generate test files.
Audit Metadata