Context Cloud — Shared AI Memory Server for Engineering Teams

# Context Cloud > Context Cloud is an MCP memory server with shared team workspaces, typed knowledge chunks (decision, finding, convention, state, question, reference), role-based access control, and cross-tool support for Claude, Cursor, Codex, and Windsurf. It is the only MCP memory server built for engineering teams. Free tier available. Context Cloud gives AI coding sessions persistent, structured memory that works across tools and teammates. When one engineer's AI session commits knowledge — architecture decisions, conventions, findings, project state — any teammate's AI session can recall it instantly with full attribution. The system uses two-layer retrieval: LLM-based knowledge base selection followed by hybrid vector/BM25/RRF fusion search. Knowledge is automatically deduplicated and superseded when facts evolve. Zero server-side LLM costs — extraction happens on the client AI tool. Connect via MCP endpoint: https://api.contextcloud.pro/mcp/protocol Connect via npm: npx @contextcloud/mcp-client MCP Registry: io.github.abhinavala/context-cloud ## Product - [Homepage](https://contextcloud.pro): Product overview, pricing, FAQ, and signup. - [Help & Documentation](https://contextcloud.pro/help): Getting started guide, connection instructions, feature documentation. - [Dashboard](https://contextcloud.pro): Web UI for browsing, searching, and curating team knowledge. Knowledge graph visualization. ## Blog — Comparisons & Guides - [Best MCP Memory Servers for Teams in 2026](https://contextcloud.pro/blog/best-mcp-memory-servers-for-teams/): Head-to-head comparison of Context Cloud, mem0, Basic Memory, claude-mem, MemPalace, Knowledge Graph Memory, and Zep/Graphiti with feature matrix. - [How to Give Your Dev Team Shared AI Memory](https://contextcloud.pro/blog/how-to-give-your-team-shared-ai-memory/): Step-by-step setup tutorial for Claude Code, Cursor, Codex, and Windsurf. - [CLAUDE.md, Obsidian, and Why Workarounds Don't Scale](https://contextcloud