codegate  by stacklok

AI agent security and management tool

created 8 months ago
692 stars

Top 50.1% on sourcepulse

GitHubView on GitHub
Project Summary

CodeGate is a security and management layer for AI agentic frameworks, coding assistants, and AI applications. It provides a centralized, abstracted environment for managing prompts, model configurations, and security, targeting developers who use AI for coding tasks. CodeGate aims to simplify AI model management and enhance security by preventing accidental exposure of sensitive data and ensuring AI-generated code adheres to secure practices.

How It Works

CodeGate operates as a local proxy, intercepting and processing interactions between AI coding assistants and AI models. It offers features like workspace management for isolating AI configurations, model muxing to route prompts to different AI models based on context, and automatic redaction of secrets and PII from prompts before they are sent to AI models. It also performs security analysis on AI-generated code and checks for vulnerable or outdated dependencies suggested by LLMs.

Quick Start & Requirements

  • Install/Run: docker run --name codegate -d -p 8989:8989 -p 9090:9090 -p 8990:8990 --mount type=volume,src=codegate_volume,dst=/app/codegate_volume --restart unless-stopped ghcr.io/stacklok/codegate:latest
  • Prerequisites: Docker Desktop/Engine or Podman. Supports Windows, macOS, Linux (x86_64, arm64).
  • Access Dashboard: http://localhost:9090
  • Docs: CodeGate Install and Upgrade documentation, CodeGate Dashboard documentation

Highlighted Details

  • Privacy-focused: All processing stays local, with no external data collection or telemetry.
  • Supports multiple AI coding assistants (Aider, Cline, Continue, GitHub Copilot, OpenInterpreter) and providers (Anthropic, OpenAI, Ollama, llama.cpp, vLLM).
  • Features secrets and PII redaction, dependency risk awareness, and security-centric code reviews.
  • Model muxing allows routing prompts to different AI models based on workspaces or file types.

Maintenance & Community

Licensing & Compatibility

  • Licensed under the terms specified in the LICENSE file (likely permissive, but requires verification).
  • Designed for local execution, ensuring compatibility with closed-source development workflows.

Limitations & Caveats

The README does not explicitly state the license type, requiring users to check the LICENSE file. Support for JetBrains IDEs for GitHub Copilot is listed as "coming soon."

Health Check
Last commit

1 month ago

Responsiveness

1 day

Pull Requests (30d)
0
Issues (30d)
0
Star History
56 stars in the last 90 days

Explore Similar Projects

Starred by Chip Huyen Chip Huyen(Author of AI Engineering, Designing Machine Learning Systems), Jeff Hammerbacher Jeff Hammerbacher(Cofounder of Cloudera), and
9 more.

codex by openai

0.8%
32k
Coding agent CLI tool for terminal-based chat-driven development
created 3 months ago
updated 1 day ago
Feedback? Help us improve.