MCPedia serves as a robust knowledge base server, enabling AI coding agents like Cursor, Claude, and Codex to access and manage persistent memory. It tackles the challenge of AI agents forgetting information between sessions by allowing users to teach agents coding patterns, project conventions, and debugging tricks once, which they will remember indefinitely. Agents can save, search, and retrieve knowledge entries—ranging from skills and rules to patterns and references—ensuring consistent access to critical information across different MCP-compatible tools. It features usage statistics, version tracking, and a write lock for controlled access, functioning as a personal wiki that AI agents can automatically read from and write to.
Características Principales
01Persistent knowledge storage for AI coding agents via MCP
02Comprehensive full-text search and contextual retrieval with advanced filtering
03Database-level write lock for controlled agent modifications
04Shareable knowledge base compatible with any MCP-enabled agent
050 GitHub stars
06Automatic version tracking, usage statistics, and tag management for entries
Casos de Uso
01Equipping AI agents with permanent access to coding standards and project-specific guidelines.
02Enabling AI agents to search for and retrieve relevant coding patterns, project conventions, or debugging tricks.
03Allowing AI agents to apply stored guidelines to current tasks or review code against established rules.
