stacklok/codegate

**Note:** This project is deprecated and no longer maintained. Provided a proxy layer offering secrets/PII redaction before AI model calls, automatic dependency vulnerability scanning against code context, and security-focused code review analysis. Supported workspace isolation with model muxing/routing (directing prompts to different models by type), plus a web dashboard tracking security detections and model interactions. Integrated with Aider, Cline, Continue, GitHub Copilot, and Open Interpreter against both local (Ollama, LM Studio) and hosted providers (OpenAI, Anthropic).

710 stars. No commits in the last 6 months.