agent-skills-manager
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious behavior, data exfiltration, or credential exposure was detected. The skill functions as a legitimate local development tool for managing skill paths.\n- [SAFE]: The skill implements a safety-conscious protocol in its instructions (SKILL.md), directing AI agents to request manual user execution for filesystem modifications. This prevents the agent from attempting to bypass environment sandbox restrictions directly.\n- [SAFE]: All filesystem operations, such as directory creation and symbolic linking, are performed using standard Python library modules (os, pathlib). No external downloads or network operations are involved in the skill's logic.\n- [SAFE]: The code does not contain any obfuscation, dynamic code execution (eval/exec), or persistence mechanisms. Its behavior aligns strictly with its stated purpose of managing local directory links.
Audit Metadata