The structure is as follows:
mcp-server-redis/ ├── src/ │ ├── __init__.py │ ├── main.py # Main entry point │ ├── config.py # Configuration │ ├── connection.py # Redis connection management │ ├── resources/ # Resource implementations │ │ ├── __init__.py │ │ ├── status.py # Connection status resources │ │ └── keys.py # Key-related resources │ └── tools/ # Tool implementations │ ├── __init__.py │ ├── basic.py # Basic Redis operations │ ├── lists.py # List operations │ ├── hashes.py # Hash operations │ ├── sets.py # Set operations │ └── pubsub.py # Pub/Sub operations ├── tests/ # Test directory │ └── __init__.py ├── README.md └── pyproject.toml To use this server, you'll need to:
Install the required dependencies:
[If not already installed] https://docs.astral.sh/uv/
uv venv
source .venv/bin/activate
uv sync
Refer to the .env.example file to configure your Redis connection by adding the appropriate variables.
Run the server:
python src/server.py
Resources:
- redis://status - Current connection status
- redis://info - Redis server information
- redis://keys/{pattern} - List keys matching a pattern
Tools:
- Basic operations:
- get_value - Get a value
- set_value - Set a value with optional expiry
- delete_key - Delete a key
- increment - Increment a numeric value
- List operations:
- list_push - Push values to a list
- list_range - Get range of list values
- Hash operations:
- hash_set - Set hash fields
- hash_get - Get hash fields
- Set operations:
- set_add - Add to a set
- set_members - Get set members
- Pub/Sub:
- publish_message - Publish a message
Each tool includes error handling and returns meaningful responses. The server will automatically reconnect to Redis if the connection is lost.
You can test it with the MCP Inspector: mcp dev src/server.py
Or install it in Claude Desktop: mcp install src/server.py
User: What's the current Redis connection status? LLM: Let me check the Redis connection status for you. [Reading resource redis://status] The Redis server is currently connected and running on localhost:6379, database 0. User: Can you store my name "Alice" in Redis? LLM: I'll help you store your name in Redis using the set_value tool. [Calling set_value with key="name" value="Alice"] Successfully set key 'name' I've stored your name "Alice" in Redis under the key "name". Would you like me to verify the value was stored correctly by retrieving it? User: Yes please verify it LLM: I'll retrieve the value using the get_value tool. [Calling get_value with key="name"] Alice Yes, I can confirm that your name "Alice" was successfully stored in Redis and can be retrieved correctly. The value matches exactly what we stored. This implementation provides a solid foundation for Redis integration through MCP. You can extend it further by adding more Redis commands as needed for your specific use case.