KNEZ Control App

The Challenge

Build a comprehensive control panel for an AI agent system (KNEZ) that can manage sessions, explore agent memory, monitor system health, and provide real-time interaction — all in a lightweight cross-platform desktop app.

The Approach

Built with Tauri v2 (Rust backend + React 19 frontend) for native performance with web UI flexibility. Implemented 30 feature modules as independent React components, with Rust handling PTY, database operations, filesystem access, and process management. Uses SQLite + PostgreSQL for data, MCP protocol for agent communication.

What We Built

• →Tauri v2 desktop application (Rust + React 19)
• →30 feature modules (Chat, Agent, Memory, MCP, Infrastructure, Governance, Reflection, Timeline, Skills, Models, Extraction, Drift, Voice, Playgrounds, Data Explorer, Terminal, FullViewer, and more)
• →Embedded terminal (xterm.js + portable-PTY)
• →AI chat interface with session management
• →Memory explorer with cognitive panel and drift visualization
• →MCP registry viewer (Model Context Protocol)
• →System health monitoring and process management
• →Database query interface (PostgreSQL + SQLite)
• →ReactFlow graph visualization
• →GitHub MCP server integration

Technology

Results