refactor-guidelines-contract
Pass
Audited by Gen Agent Trust Hub on Jul 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is instructional and meta-behavioral, setting standards for the agent's execution behavior during a refactoring pipeline. Analysis of all skill files revealed no malicious patterns, prompt injections, or data exfiltration attempts.
- [COMMAND_EXECUTION]: The skill includes a local Node.js script,
scripts/checklist.mjs, which provides the contract clauses in JSON format. The script is minimal, contains no external dependencies, and performs no dangerous system or network operations. - [DATA_EXPOSURE]: No hardcoded credentials or sensitive file paths were identified. The skill encourages best practices by instructing the agent to verify API existence through local grep commands and checking local manifest files.
Audit Metadata