Connect your own AI

The iRIC knowledge base (manuals, tutorials, examples, FAQ) is available as an MCP server. Connect your Claude or a local AI and it can answer how-to questions with sources.

Your tokens

How to connect

Endpoint: https://software-dev.i-ric.org/mcp

Claude (claude.ai / Claude Desktop)

Settings → Connectors → "Add custom connector", register the URL below. When asked for authentication, enter your token as a Bearer token.

Claude Code

Run this in a terminal.

claude mcp add --transport http iric https://software-dev.i-ric.org/mcp --header "Authorization: Bearer <YOUR_TOKEN>"

Ollama (local LLM)

Ollama's built-in integration runs Claude Code on an Ollama model; connect from there. No Anthropic subscription is needed. Use a model that supports tool calling (e.g. qwen3:4b). The first time, run the claude mcp add command above inside the launched Claude Code.

ollama pull qwen3:4b
ollama launch claude --model qwen3:4b

On a PC without a GPU, 12B-class models take minutes per turn. A model around 4B, or an NVIDIA GPU, is recommended.

Other local AI (LM Studio and other MCP-capable clients)

Any MCP-capable client can register the URL and token as a Streamable HTTP server. For clients that only speak stdio, bridge it with:

npx mcp-remote https://software-dev.i-ric.org/mcp --header "Authorization: Bearer <YOUR_TOKEN>"

Tokens are for your own use; do not share them. The limit is 120 calls per minute. Results are excerpts with sources; distribution files are not included.