universal-planner
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill follows a legitimate and well-structured architectural pattern for managing software development workflows without malicious intent.
- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute implementation and verification tasks. This is an essential feature of its EXECUTE mode and is governed by structured planning documents and graduation gates.
- [PROMPT_INJECTION]: The skill includes robust internal rules regarding mode boundaries and the prioritization of project-specific conventions, which act as safeguards against unintended behavioral overrides. No malicious jailbreak or bypass patterns were detected.
- [DATA_EXFILTRATION]: The skill reads local project files and configuration (e.g., AGENTS.md) to inform its planning logic. No patterns were found indicating that this information is transmitted to untrusted external domains or that sensitive credentials are hardcoded.
Audit Metadata