lint-build-loop
Pass
Audited by Gen Agent Trust Hub on Feb 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes sequential shell commands (
npm run lint && npm run build) to identify and iterate on code issues. - [PROMPT_INJECTION]: The workflow is susceptible to indirect prompt injection (Category 8) because it processes untrusted tool output as instruction feedback.
- Ingestion points: The agent reads error output from development scripts in
SKILL.mdto determine how to fix the code. - Boundary markers: Absent; there are no delimiters or instructions provided to the agent to treat tool output as non-instructional data.
- Capability inventory: The skill allows for direct shell command execution and file modification within the project environment.
- Sanitization: No validation or sanitization of the tool output is performed before the agent interprets it for the next iteration.
Audit Metadata