skills/rkz91/coco/agent-lightning/Gen Agent Trust Hub

agent-lightning

Pass

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [SAFE]: No security issues detected.
  • [EXTERNAL_DOWNLOADS]: Fetches the 'agentlightning' package from the Python Package Index (PyPI) and official test repositories for nightly builds. These sources are standard for the tool's installation.
  • [COMMAND_EXECUTION]: Includes standard shell commands for package management and server orchestration during development and training.
  • [DATA_EXFILTRATION]: Utilizes network listeners on local interfaces for inter-component communication, which is a documented and standard feature of the training framework.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 23, 2026, 06:23 AM
Security Audit — agent-trust-hub — agent-lightning