Recall
A remote MCP server for personal memory, enabling AI tools like Claude Code to save and search memories with user permission, built on Appwrite.
README
Recall: a memory MCP server on Appwrite Functions
Recall is a small personal-memory app built on Appwrite. It ships a remote MCP server, hosted as an Appwrite Function, so AI tools like Claude Code can save and search a user's memories with the user's permission.
Memories are stored in an Appwrite VectorsDB database: each memory is embedded with the all-minilm model and retrieved by meaning with a cosine-similarity query. Auth runs on the project's OAuth2 server: AI clients register themselves, users approve access on a consent screen Recall hosts, and the MCP server verifies the tokens the project issues. No API keys are handed out, and users can disconnect a tool at any time.
Companion repository for the Appwrite blog tutorial "Build a memory MCP server on Appwrite".
Layout
apps/recall The Recall web app (TanStack Start): memory dashboard,
email sign-in, the OAuth2 consent screen, and a
connected apps page.
functions/memory-mcp The MCP server, deployed as an Appwrite Function.
Setup
-
Create an Appwrite project.
-
In the Console, create a VectorsDB database with ID
recall, then a collection with IDmemoriesusing theall-minilm(384) embedding model, and add anhnsw_cosineindex on theembeddingsfield. -
Create an API key with the scopes
sessions.write,users.read,apps.read,databases.read,collections.read,collections.write,documents.read,documents.write, andembeddings.write. -
Copy
.env.exampleto.envwith your endpoint, project ID, the API key, and a random session secret. -
In the Console, open Auth > OAuth2 server, enable the server, set the authorization URL to
http://localhost:4200/oauth/consent, and add thememories.readandmemories.writescopes. -
Deploy the web app and the MCP server (uses the Appwrite CLI), then set the site's variables in the Console (same values as
.env):pnpm install appwrite login appwrite push site --site-id recall appwrite push function --function-id memory-mcp --activate -
Update the OAuth2 server's authorization URL to
https://<SITE_DOMAIN>/oauth/consent. -
Connect from Claude Code, using the function's domain from the deploy output:
claude mcp add --transport http recall https://<FUNCTION_DOMAIN>/mcpRun
/mcpinside Claude Code and pick Authenticate. Sign in as a Recall user, approve access, and ask Claude to remember something.
Live demo
The deployed result of this repository: the app at recall.appwrite.network and the MCP server at https://recall-mcp.appwrite.network/mcp. Create an account, connect Claude Code with the command above, and ask it to remember things.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.