teach

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a structured workflow for multi-session teaching by managing state in a dedicated project directory (.agents/projects/<project>/). All file operations are scoped to this specific workspace, which is consistent with the skill's stated purpose of persistent project management.
  • [SAFE]: The skill provides explicit formatting rules for its metadata files (Mission, Resources, Glossary, Learning Records) and encourages the use of high-trust primary sources, which helps maintain data integrity and provides a clear structure for the agent's operations.
  • [SAFE]: No malicious patterns such as prompt injection, data exfiltration, or unauthorized command execution were detected. The skill's focus is on organizational and pedagogical instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 01:25 PM
Security Audit — agent-trust-hub — teach