MemoryRouter MCP Server

About MemoryRouter

MemoryRouter is a hosted MCP memory server. Connect it once and an agent gets a persistent, user-scoped memory that outlives the session and the model.

The same vault follows the user across ChatGPT, Claude, Cowork, Claude Code, Codex, OpenClaw and DeepSeek Harness. Claude opens with what ChatGPT learned about the project this morning, and a coding agent picks up a decision made two weeks ago.

Tools exposed:
search_memories - semantic recall over the vault, with optional tier and importance filters
date_search_memories - pull memories from a specific window
store_memory - save something worth keeping
consolidate_memories and commit_reflections - fold raw memories into higher-level reflections
inspect_memory - trace a reflection back to the memories underneath it
delete_memories and forget_all_memories - hard delete at vault level
memory_status - what is in the vault right now

Memory is tiered. Raw verbatim memories sit at the bottom, reflections of specific events above them, and a consolidated layer on top holding the highest-level reflections. An agent can start at a high-level reflection and drill down to the exact wording.

Isolation is per vault. Vaults are scoped at the retrieval layer with no shared index, one API key reaches one vault, and there are no cross-tenant retrieval paths. TLS 1.2+ in transit, AES-256 at rest, API keys hashed before storage. Deletion is a hard delete at the vault level rather than a flag.

Typical recall is 300 to 500 ms for a vault under about 15,000 vectors. Vaults several times that size answer in the low seconds.

Works with any MCP client that speaks Streamable HTTP. OAuth scopes a connection to one selected vault.

MCP setup guide: https://memoryrouter.ai/mcp

14 days free, $20 a month after, no card needed to start.

MemoryRouter tools (10)

Read from the live server on Nov 8, 2026. These are the tools your AI assistant can call once connected.

search_memories

Use this before responding to each new user message, to retrieve context relevant to the message and ongoing conversation, even when the user does not explicitly ask you to remember. Form a concise semantic query using the people, projects,

date_search_memories

Use this to retrieve memories from a specific time window, for questions like 'what happened last week' or 'most important things this month or lately'. Tier guide: tier 3 holds the highest-level consolidated reflections (identity, principl

store_memory

Use this after each user message that contains anything worth knowing later, even if the user does not say 'remember this'. Default to saving: a missed memory costs far more than a redundant one, so when in doubt, save. Each of these is a t

memory_status

Use this when the user asks whether MemoryRouter is connected, which opaque vault binding is active, or how many memories/tokens the connected vault contains. Does not reveal the memory key.

forget_all_memories

Use this only after the user explicitly requests a full-vault deletion, receives a permanent whole-vault warning, and provides the exact confirmation phrase in a later turn. Permanently deletes every memory in the OAuth-selected vault. This

consolidate_memories

Use this when the user asks to consolidate, reflect on, or organize their memories, or when memory_status or a search result reports that consolidation is available. Checks out a batch of unconsolidated memories from the connected vault und

commit_reflections

Use this to complete a consolidation started with consolidate_memories. Submits the reflection entries you wrote for the checked-out batch. Each entry is standalone prose with an importance rating from 1 to 10. The server stores the reflect

inspect_memory

Use this when exact details or provenance matter: when the user asks why the assistant believes something, or wants to see the source memories behind a reflection. Given a memory id from a search result, returns the lower-tier memories that

delete_memories

Use this only when the user explicitly asks to delete specific memories and confirms after seeing what will be deleted. Permanently deletes the identified memories from the connected vault by their ids from search or inspect results. Deleti

search

Use this legacy compatibility alias when a client explicitly requests a tool named search. It searches the OAuth-selected MemoryRouter vault without modifying it. For normal ChatGPT recall, prefer search_memories; this tool does not provide

How to connect MemoryRouter to Claude, Cursor or VS Code

MemoryRouter runs as a hosted remote MCP server at https://mcp.memoryrouter.ai/mcp. Add that URL to your MCP client's config (Claude Code, Cursor, Windsurf or VS Code) and the tools appear immediately, no install required.

{
  "mcpServers": {
    "memoryrouter": {
      "url": "https://mcp.memoryrouter.ai/mcp"
    }
  }
}

Claude Code: claude mcp add --transport http memoryrouter https://mcp.memoryrouter.ai/mcp. VS Code's .vscode/mcp.json uses a servers key instead of mcpServers (see how). Servers that require OAuth or an API key will prompt on first use.