brain-projects

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill retrieves project data using dedicated MCP tools (mcp__graph-brain__list-projects, mcp__graph-brain__get-project, mcp__graph-brain__tech-overview), which is a restricted and intended method for data access.
  • [SAFE]: File operations are limited to recording session metadata and activity summaries into a local JSON log file (.brain/activity-log.json). This is a standard logging practice and does not involve sensitive system files.
  • [SAFE]: No network exfiltration, remote code execution, or prompt injection patterns were detected. The skill operates entirely within the local context and predefined toolset.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 06:37 PM