mcp-skills-server
Verified Safeby tomoharu-hayashi
Overview
The MCP Brain Server acts as a long-term memory and knowledge base for AI agents, allowing them to learn from past experiences, store new knowledge, and retrieve relevant information via semantic search, mimicking human learning processes.
Installation
uvx --from git+https://github.com/tomoharu-hayashi/mcp-server-brain.git mcp-brainEnvironment Variables
- MCP_BRAIN_DIR
Security Notes
The server primarily operates locally, interacting with the file system and Git. Input validation for knowledge names (kebab-case pattern) helps prevent directory traversal. `pyyaml.safe_load` is used for parsing knowledge files, mitigating YAML-related vulnerabilities. `pickle.load` is used for the embedding cache; while a known risk, it's mitigated here as the cache file is internally generated and its integrity is checked against a hash of the knowledge directory content. macOS-specific dialogs use AppleScript with input escaping (`_escape_applescript`) to prevent injection. Git operations (commit, push) are handled with `gitpython`, including mechanisms to prevent or resolve conflicts (rebase). The server requires a trusted Git repository and user interaction for `create`/`update` tools, making it generally safe for its intended use within a developer's environment.
Similar Servers
zen-mcp-server
A server for coordinating and managing AI agents, likely for simulations or complex task execution, leveraging Claude LLMs.
llms
A centralized configuration and documentation management system for LLMs, providing tools for building skills, commands, agents, prompts, and managing MCP servers across multiple LLM providers.
simplenote-mcp-server
Integrates Simplenote with Claude Desktop as a memory backend or content source via the Model Context Protocol (MCP), enabling AI interaction with user notes.
tiger-skills-mcp-server
This server emulates Anthropic's skill format, enabling LLMs to utilize specialized knowledge and tools via the Model Context Protocol (MCP).