Kenaz documentation

Welcome to Kenaz, by Kameas AI — a local-first platform for AI-assisted work. Your prompts, keys, and files stay on your machine; the cloud is optional and only does what a team explicitly turns on.

Kenaz is three products:

• Kenaz Harness — open source. The local AI agent app you actually click. Bring your own provider keys (Anthropic, OpenAI, Azure OpenAI, Google Gemini, OpenRouter, Bedrock, Ollama, or any OpenAI-compatible endpoint), run tools and MCP servers under a permission prompt, manage sessions, context, and memory, and get a tamper-evident local audit log. Runs on its own or inside the Workbench.
• Kenaz Workbench — open source. The complete environment that runs the Harness inside an isolated sandbox (Tart on macOS, WSL2 on Windows). This is the recommended way to run Kenaz — the Harness is built in.
• Kenaz Fleet — hosted by Kameas AI. The optional cloud control plane for teams and organizations: single sign-on, org/team management, roles and plans, and pushing shared configuration, skills, and tools down to every member's Harness. The Harness and Workbench are fully usable without it; Fleet adds team management on top.

Open source vs. hosted

The Harness and Workbench are open source — you can read the code, build them yourself, and run them with no account. Fleet is the hosted, closed-source service Kameas AI operates for teams. Signing a Harness in to Fleet is always opt-in, and what Fleet can push to a Harness is bounded by the member's role and plan.

Where to start

• Just want to use it? Download the Workbench — the Harness is included.
• Running the Harness on its own? Start with the Harness quickstart.
• Setting up a team? Read the Fleet overview.

Getting help

• File an issue against the relevant repo on GitHub (kenaz-harness, kenaz-workbench).
• For private or security questions, email security@kameas.ai.