agentmemory-mcp-shim
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@agentmemory-mcp-shimremember that the project uses React 18 and Vite"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
agentmemory-mcp-shim
Coding agents and other local MCP hosts spawn this process. It forwards tools/list and tools/call to AgentMemory REST and returns the server payload as-is, including structuredContent. There is no local memory database and no silent fallback.
Why this exists
npx @agentmemory/mcp runs upstream @agentmemory/agentmemory, which drops structuredContent on the stdio path. Cursor then rejects every tool call with -32600 once the tool advertises outputSchema.
This shim is the Cursor-side counterpart of agentmemory-mcp-gateway: same two REST endpoints, fail loud when the server is missing.
Related MCP server: GroundMemory
Required env
Name | Purpose |
| AgentMemory origin, e.g. |
| Bearer secret for |
Missing either one exits with status 1. Optional: AGENTMEMORY_TIMEOUT_MS (default 10000).
Setup
This is a complete mcp.json. Paste it into the host config (Cursor uses ~/.cursor/mcp.json) and replace the env values:
{
"mcpServers": {
"agentmemory": {
"command": "npx",
"args": ["-y", "@martindzejky/agentmemory-mcp-shim"],
"env": {
"AGENTMEMORY_URL": "https://memory.example.com",
"AGENTMEMORY_SECRET": "replace-me"
}
}
}
}Wire protocol
GET ${AGENTMEMORY_URL}/agentmemory/mcp/toolsPOST ${AGENTMEMORY_URL}/agentmemory/mcp/callwith{ "name", "arguments" }Pass through
content,structuredContent,isError, and_meta
Upstream errors become MCP tool errors (isError: true). This process never opens an InMemoryKV or writes observations locally.
This server cannot be deployed
Maintenance
Related MCP Connectors
The Remote MCP server acts as a standardized bridge between LLM applications (like Claude, ChatGPT, and Cursor) and external services, enabling AI agents to access external tools and resources. Its primary capability is providing a centralized search tool to discover other MCP servers and their respective tools. Unlike local implementations, it runs remotely with OAuth authentication and permission controls for security.
Cross-tool persistent memory and context for AI assistants over MCP.
- memnodeOAuthdev.memnode
Persistent, inspectable memory for AI agents with lineage, correction, and a hosted MCP endpoint.
Remote MCP server for supportsheep: run AI interviews and manage support content for your blog.
Related MCP Servers
AlicenseNot gradedqualityDmaintenanceMCP server providing managed persistent memory for AI agents. Read and write structured state across sessions, tools, and restarts at 1000+ requests per second, with no infrastructure to self-host or operate.2Apache 2.0- AlicenseNot gradedqualityCmaintenanceAn MCP-native, local-first memory server that gives AI agents persistent, structured memory across sessions and tools, enabling them to maintain identity and context without reconfiguration.3MIT
- AlicenseNot gradedqualityCmaintenanceEnables MCP-compatible coding agents to read and write persistent, Markdown-backed memory via the EverOS REST API.Apache 2.0
- FlicenseNot gradedqualityBmaintenanceEnables remote MCP clients to securely access AgentMemory tools like memory_recall, memory_smart_search, and memory_save through OAuth 2.1 and Streamable HTTP, while keeping the AgentMemory backend secret private.-