Remind

v1.0.0Knowledge & Memorystable

A memory layer for AI Agents

agentic-aiagentsaicodingcontext-engineering
Share:
79
Stars
0
Downloads
0
Weekly
0/5

What is Remind?

Remind is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to memory layer for ai agents

A memory layer for AI Agents

This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • A memory layer for AI Agents

Use Cases

Memory layer for AI agents
Long-term memory management
Context engineering for coding
sandst1

Maintainer

LicenseApache-2.0
Languagepython
Versionv1.0.0
UpdatedMay 20, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx remind

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Remind

Remind is a generalization-capable long-term memory layer for LLM agents that goes beyond simple RAG by extracting abstract concepts from specific experiences — mimicking how human memory consolidates events into reusable knowledge. Unlike verbatim text storage, it uses spreading activation retrieval, entity graph extraction, outcome tracking, and memory decay to give agents persistent, evolving understanding across sessions. AI engineers and developers use it to give coding agents, chat assistants, and other LLM systems the ability to remember project context, learned patterns, and past decisions between conversations.

Prerequisites

  • Python 3.11 or higher installed
  • An Anthropic API key (sk-ant-...) for the LLM provider used to extract generalizations
  • An OpenAI API key (sk-...) for the embedding provider (default is OpenAI embeddings)
  • Optional: a PostgreSQL or MySQL database for persistent storage (SQLite used by default)
  • An MCP-compatible client such as Claude Desktop or Claude Code
1

Install the remind-mcp package

Install the base package from PyPI. Add optional extras for PostgreSQL or MySQL backends, or for cross-encoder reranking support.

# Base install
pip install remind-mcp

# With PostgreSQL support
pip install "remind-mcp[postgres]"

# With cross-encoder reranking
pip install "remind-mcp[rerank]"
2

Create the provider configuration file

Create the config directory and write the provider config file. This tells Remind which LLM to use for concept extraction and which embedding model to use for semantic retrieval.

mkdir -p ~/.remind
cat > ~/.remind/remind.config.json << 'EOF'
{
  "llm_provider": "anthropic",
  "embedding_provider": "openai",
  "anthropic": { "api_key": "sk-ant-your-anthropic-key" },
  "openai": { "api_key": "sk-your-openai-key" }
}
EOF
3

Optional: configure a persistent database backend

By default Remind uses SQLite. For production or multi-agent setups, configure PostgreSQL by adding the db_url to the config file or setting the REMIND_DB_URL environment variable.

# Add to ~/.remind/remind.config.json:
# "db_url": "postgresql+psycopg://user:password@localhost:5432/remind"

# Or via environment variable:
export REMIND_DB_URL="postgresql+psycopg://user:password@localhost:5432/remind"
4

Start the Remind MCP server

Launch the server on the default port 8765. The server runs as an SSE-based MCP endpoint that clients connect to via URL.

remind-mcp --port 8765
5

Configure your MCP client to connect to Remind

Add the Remind server to your Claude Desktop or other MCP client config. The connection uses an SSE URL rather than a subprocess command. Use the db query parameter to namespace memories per project.

Remind Examples

Client configuration

Remind connects via SSE URL rather than a command. Add this to your claude_desktop_config.json. The 'db' parameter creates a named memory namespace for your project.

{
  "mcpServers": {
    "remind": {
      "url": "http://127.0.0.1:8765/sse?db=my-project"
    }
  }
}

Prompts to try

Once connected, Remind gives your AI assistant persistent memory across sessions.

- "Remember that we decided to use PostgreSQL for this project's database"
- "What do you remember about our authentication approach?"
- "Store the outcome: the refactor of the payment module reduced latency by 40%"
- "What patterns have you learned about how I like to structure React components?"

Troubleshooting Remind

MCP client cannot connect to the Remind server at localhost:8765

Ensure the Remind server is running before starting the MCP client: run 'remind-mcp --port 8765' in a terminal and leave it running. The server must be started manually and kept alive — it is not auto-launched from the config URL.

Concept extraction fails with Anthropic API errors

Verify the api_key in ~/.remind/remind.config.json is a valid Anthropic key starting with 'sk-ant-'. Check that the key has sufficient credits and is not rate-limited. You can test it independently with 'curl https://api.anthropic.com/v1/messages -H "x-api-key: sk-ant-..."'.

Memories are not persisting between server restarts

By default Remind uses an in-memory SQLite database that is wiped on restart. Set a file-based SQLite path in the config: 'db_url': 'sqlite:///home/user/.remind/memories.db', or configure a PostgreSQL database for durable persistent storage.

Frequently Asked Questions about Remind

What is Remind?

Remind is a Model Context Protocol (MCP) server that memory layer for ai agents It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Remind?

Follow the installation instructions on the Remind GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Remind?

Remind works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Remind free to use?

Yes, Remind is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.

Browse More Knowledge & Memory MCP Servers

Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "remind": { "command": "npx", "args": ["-y", "remind"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Remind?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides