loop-runner
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands for its core functionality of verifying and revising code. It accepts command strings via parameters like --verify and --revise.
- [PROMPT_INJECTION]: An indirect prompt injection surface exists via the workspace state and the episodic-memory wiki page. 1. Ingestion points: workspace files (src/**) and the output of the revision command. 2. Boundary markers: none mentioned. 3. Capability inventory: arbitrary shell command execution and local file writes. 4. Sanitization: none mentioned.
- [SAFE]: The skill implements deterministic guards including iteration caps, wall-clock budgets, and state-hash progress detection to mitigate the risk of infinite loops or uncontrolled agent behavior. No malicious patterns or unauthorized data exfiltration were detected.
Audit Metadata