chatkit-botbuilder

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides architecture and implementation guides for building secure chatbots, focusing on multi-user isolation and standard authentication patterns.
  • [SAFE]: Implements a comprehensive 'Three-Level Isolation Strategy' (Middleware, Tool Wrapper, and Database filtering) to ensure that users can only access their own conversation data and tasks.
  • [SAFE]: All dependencies and external references are to well-known, trusted technology providers such as OpenAI, FastAPI, and Vercel.
  • [INDIRECT_PROMPT_INJECTION]: The skill documents an architecture for processing user input, which inherently has an indirect prompt injection surface. The documentation includes mitigation strategies such as user context boundaries and parameter sanitization. (Ingestion: ChatKit endpoint in references/backend_architecture.md; Boundaries: Mentioned in references/user_isolation.md; Capabilities: Database and tool execution via MCP in references/mcp_wrapper_guide.md; Sanitization: Documented as a practice in references/user_isolation.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 04:33 PM
Security Audit — agent-trust-hub — chatkit-botbuilder