Kestrel releases open-source sovereign AI agent framework with user-owned identity, persistent memory, and constitutional governance

Kestrel is a production-ready framework for creating autonomous AI agents where each agent is owned by its user, governed by immutable principles, and able to remember across conversations. Core features include cryptographic identity (DID), SQLite-backed persistent memory with full-text search and RAG, and constitutional governance enforced above the LLM.

The base installation (`pip install kestrel-sovereign`) provides a complete working agent with identity, memory, constitution, privacy modes, multi-LLM support, voice (Piper TTS + FasterWhisper STT), local sandboxed compute, and a Cloud Run deployment path. Cloud providers (RunPod, Vast.ai), specialized integrations (MCP, GitHub App, wallet), and proprietary training adapters are installable add-ons registered via Python entry points.

Agents run locally on configurable ports (default 8888) and expose a web interface (Sovereign Console) with 8 tabs for identity, chat, constitution audit, memory browsing, task monitoring, sovereignty management, resource usage, and security controls. An OpenAI-compatible API at `/v1/chat/completions` allows connection to third-party clients like Open WebUI.