Getting Started

Prerequisites

Before you begin, make sure you have:

• Docker (or a Node 20+ runtime)
• PostgreSQL 15+ with the pgvector extension
• A GitHub App (see GitHub connector setup)
• An Anthropic API key for LLM processing
• An OpenAI API key for embeddings

Quickstart with Docker

1. Clone the repository

git clone https://github.com/your-org/mneme.git
cd mneme

2. Create an environment file

cp .env.example .env

Edit .env with your credentials:

DATABASE_URL=postgresql://user:password@localhost:5432/mneme

GITHUB_APP_ID=your-app-id
GITHUB_PRIVATE_KEY="-----BEGIN RSA PRIVATE KEY-----\n...\n-----END RSA PRIVATE KEY-----"
GITHUB_WEBHOOK_SECRET=your-webhook-secret

ANTHROPIC_API_KEY=sk-ant-...
OPENAI_API_KEY=sk-...

API_TOKEN=your-api-token

3. Start the services

docker compose up -d

This starts Mneme along with PostgreSQL. No Redis needed — pg-boss uses Postgres for job queues.

4. Run database migrations

pnpm db:migrate

5. Install the GitHub App

Go to your GitHub App settings and install it on the repositories you want Mneme to monitor. Point the webhook URL to:

https://your-mneme-host/webhooks/github

6. Verify

Send a test webhook from GitHub App settings or open a pull request on an installed repo. Check the Mneme logs:

docker compose logs -f mneme

You should see the webhook received and processing begin.

What's Next

• Installation — full deployment guide for Docker and Fly.io
• Configuration — customize behavior via the dashboard settings
• Connectors — add Slack and Linear signals