Getting Started

Installation

pip

pip install remind-mcp

pipx

pipx install remind-mcp

uv

# No install needed — run directly
uvx --from remind-mcp remind --help

Docker

git clone https://github.com/sandst1/remind.git
cd remind
cp .env.example .env   # Edit with your API keys
docker compose up -d

Configure a provider

Remind needs an LLM provider (for consolidation) and an embedding provider (for retrieval). The simplest setup:

Create ~/.remind/remind.config.json:

{
  "llm_provider": "anthropic",
  "embedding_provider": "openai",
  "anthropic": {
    "api_key": "sk-ant-..."
  },
  "openai": {
    "api_key": "sk-..."
  }
}

Or use environment variables:

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

See Configuration for all provider options including Azure OpenAI and Ollama (fully local).

Your first memory

# Store some experiences
remind remember "This project uses React with TypeScript"
remind remember "We chose PostgreSQL for the database" -t decision
remind remember "All API routes need authentication" -t spec

# Consolidate into concepts
remind consolidate

# Retrieve
remind recall "What tech stack are we using?"

That's it. Episodes go in, consolidation runs, generalized concepts come out.

Next steps

• Skills + CLI — The recommended integration path. Use Remind as a memory primitive inside agent skills.
• MCP Server — Run Remind as a centralized tool server for IDE agents.
• Core Concepts — Understand episodes, consolidation, and the concept graph.
• Examples — See Remind used for project memory, sparring, and research ingestion.