Open-source platform for managing AI agent teams with built-in governance, budget enforcement, and compliance. Model-agnostic, air-gap capable, self-hosted.
SIDJUA runs where your team already works. No GPU required.
V1 runs on Docker with native desktop frontend (macOS, Windows 10/11, Ubuntu — Tauri-based). V2 migrates the server to Rust — targeted end of 2026.
5-step enforcement pipeline validates every agent action before execution. Policy violations are blocked before execution — with full audit trail.
Per-agent, per-division, per-task spending limits with real-time cost tracking. Never get a surprise API bill again.
Works with any LLM: Anthropic, OpenAI, Google, Groq, DeepSeek, local models via Ollama. Switch providers without code changes.
Three commands to a working system. Built-in AI guide helps with setup. Free tier available via Cloudflare Workers AI.
Runs entirely on your infrastructure. No cloud dependency. Deploy on-premises, in your VPC, or fully offline with local models.
Encrypted secrets manager with RBAC. AES-256-GCM + argon2 key derivation. No external vault required.
git clone https://github.com/GoetzKohlberg/sidjua.git cd sidjua docker compose up -d docker exec -it sidjua sidjua init docker exec -it sidjua sidjua chat guide
Requires Docker and Docker Compose. See full instructions on GitHub.
21 implementation phases. Governance framework, runtime, CLI, REST API.
External security audit, budget enforcement, secrets management, Docker packaging.
Deep security hardening based on external audit. Enhanced governance enforcement, strengthened input validation, improved system resilience.
Public release with Tauri desktop frontend. Custom action types. Pilot customer program for regulated industries.
Rust server rewrite: native binary for macOS, Windows, Linux — no Docker, no Node.js. Built-in management GUI.
Multi-location deployment, compliance certifications, vertical industry editions.
Enterprise inquiries, pilot program, or partnership?
contact@sidjua.com