tinybrain

TinyBrain is a security-focused MCP server that provides intelligent memory storage and retrieval capabilities for security professionals, red teams, and AI assistants. It exposes a 40-tool MCP suite that covers core memory operations, session and task management, advanced memory features, security templates, memory lifecycle cleanup, and AI-enhanced search. You can store, retrieve, search, and relate memory entries, manage assessment sessions and tasks, create relationships and context snapshots, and perform semantic and embedding-based searches to quickly surface relevant information. Use the MCP endpoint to issue tool calls, fetch results, and reason about complex security investigations with context-aware summaries and real-time updates.

To use TinyBrain, deploy the Docker image and connect your MCP client to the running container. The toolset is designed to interoperate with LLMs and automation scripts, enabling semantic queries, memory provenance tracking, and structured outputs that map to MITRE ATT&CK concepts, intelligence patterns, and security references. Tools are grouped into core memory operations, session management, advanced memory features, security templates, and lifecycle/cleanup tasks, allowing you to perform end-to-end security assessments and intelligence workflows within a cohesive memory-enabled environment.

Prerequisites:

• Docker installed and running on your host
• Basic familiarity with MCP tool invocation patterns

Installation steps:

1. Pull the TinyBrain MCP Docker image: docker pull rainmana/tinybrain:latest

2. Run the container in interactive mode, exposing the MCP endpoint as needed (adjust ports as required by your setup): docker run -it --rm --name tinybrain-mcp rainmana/tinybrain:latest

3. Verify the MCP server is reachable from your MCP client and begin issuing MCP tool calls. If your client requires a specific host/port, expose the necessary ports in the docker run command (e.g., -p 8080:8080).

4. (Optional) Persist data by mounting a volume: docker run -it --rm -v /path/to/tinybrain/data:/data rainmana/tinybrain:latest

Prerequisites overview:

• Docker must be installed and running
• Access to the TinyBrain Docker image (rainmana/tinybrain)
• A compatible MCP client to issue calls to the server

Tips and caveats:

• Environment variables for configuration are not required for a default Docker run, but you can extend the container with volumes for persistence and with environment variables if the image supports them (check the image docs).
• If you encounter memory or performance issues, consider using a persistent data volume and ensuring adequate CPU/memory allocation for the container.
• The 40 MCP tools cover a broad range of operations; refer to the complete tool list in the README to discover the exact syntax for each tool call.
• When integrating with MITRE ATT&CK data, ensure your data sources are updated regularly to reflect the latest techniques and mappings.
• For production deployments, implement proper access control and logging to monitor tool usage and memory access patterns.