Quick Start

    Get Attach Gateway running in under 60 seconds. Add secure authentication and memory to any LLM engine.

    No registration requiredWorks with Ollama, vLLM, and moreLocal-first setup

    Prerequisites

    Make sure you have these installed before starting

    Python 3.12+
    Docker installed
    Auth0 account or DID token
    1

    Install Attach Gateway

    Get started with a simple pip install

    pip install attach-dev
    2

    Start Memory Backend

    Launch Weaviate in Docker for context storage

    docker run --rm -d -p 6666:8080 \
  -e AUTHENTICATION_ANONYMOUS_ACCESS_ENABLED=true \
  semitechnologies/weaviate:1.30.5
    3

    Configure Environment

    Set up your authentication and memory settings

    export JWT="<your-auth-token>"
export OIDC_ISSUER=https://your-domain.auth0.com
export OIDC_AUD=ollama-local
export MEM_BACKEND=weaviate
export WEAVIATE_URL=http://127.0.0.1:6666
    4

    Launch Gateway

    Start the gateway and make your first authenticated request

    attach-gateway --port 8080 &

curl -H "Authorization: Bearer $JWT" \
     -d '{"model":"tinyllama","messages":[{"role":"user","content":"hello"}]}' \
    http://localhost:8080/api/chat | jq .

    What You Get

    Once running, Attach Gateway provides these capabilities out of the box

    OIDC/DID-JWT authentication
    Pluggable memory backends
    Agent-to-agent protocols
    Built-in observability
    Token usage tracking
    Zero-trust security

    What's Next?

    Now that you have Attach Gateway running, explore the documentation or check out the source code on GitHub.