Skip to content

Getting Started

This tutorial walks you through installing Distill, connecting it to Claude Code, and storing your first team memory.

Prerequisites

  • Python 3.11+
  • Ollama installed and running
  • Claude Code CLI

Step 1: Install Distill

pip install distill-mcp

Or from source:

git clone https://github.com/5queezer/distill.git
cd distill
uv sync

Step 2: Pull the required Ollama models

Distill needs two models — one for distillation (turning your raw text into anonymous facts) and one for embeddings (enabling semantic search):

ollama pull gemma3:4b          # distillation
ollama pull nomic-embed-text   # embeddings (768-dim vectors)

Step 3: Register with Claude Code

claude mcp add distill -- python -m distill_mcp

Or if running from source:

claude mcp add distill -- uv run python -m distill_mcp

Step 4: Store your first memory

Open Claude Code and say:

You:    "Remember that we chose gRPC over REST for inter-service
         communication because of streaming support and type safety."

Claude will:

  1. Send the raw text to your local Ollama for distillation
  2. Show you the distilled preview (e.g., "gRPC was chosen over REST for inter-service communication due to streaming support and type safety.")
  3. Wait for your approval
  4. Store the approved fact in the team database

Step 5: Search your memories

You:    "How do our services communicate?"

Claude: [searches team memory]
        Based on your team's knowledge base, inter-service communication
        uses gRPC, chosen for streaming support and type safety.

What you've learned

  • How to install Distill and its Ollama dependencies
  • How to register the MCP server with Claude Code
  • The two-phase remember flow: distill → preview → confirm
  • How search retrieves stored team knowledge

Next steps