Model Context Protocol Memory Layer
Connect Once.
Remember Everywhere.
A long-term memory engine for AI coding assistants. Automatically persists project architecture, database schemas, and coding preferences across Cursor, Claude Desktop, Antigravity, and Windsurf.
How It Works
01
Connect via MCP
Add Tanvelo to Cursor or Claude Desktop config in 1 line. No custom plugins required.
02
Autonomous Ingestion
When you discuss tech decisions or preferences, Tanvelo saves them in the background via save_memory.
03
Pre-Flight Context
AI assistants query get_context before generating code. Zero context amnesia.
Quick AI Setup Prompt
Give this prompt to your AI assistant to automatically connect to Tanvelo MCP.
💡 Works with Cursor, Claude Desktop, Google Antigravity, Windsurf, Claude Code, Cline, and any MCP-compatible AI.
Exposed MCP Tools
Standard tools available to AI assistants over the MCP protocol.
| Tool Name | Description | Action |
|---|---|---|
| get_context | Pre-flight context retrieval before code generation. | Pre-flight |
| save_memory | Autonomous background storage of facts & preferences. | Storage |
| search_memory | Semantic vector search + hybrid ranking. | Search |
| update_memory | In-place semantic update for modified architectural decisions. | Update |
| forget_memory | Delete or invalidate memory by ID or natural language query. | Delete |
| list_memories | Paginated list with category and project filters. | List |
| export_memories | Export memory vault to Markdown or JSON. | Export |