One memory for ChatGPT, and every tool you will use
Connect Hypermemory to ChatGPT via MCP, or use the CLI where you automate. Your chats, agents, and Claude sessions all draw from one graph you own.
Why This Works
Hypermemory is accessed via MCP and the CLI, the same memory graph either way. In ChatGPT, add the MCP connector and skill for hm_ tools in conversation. For scripts and other stacks, use the CLI with the same API key and graph ID. Change tools without migrating embeddings or re-uploading docs.
Quick Setup
MCP URL: https://api.hypermemory.io/mcp
- Copy the MCP URL below and paste it as a custom connector in ChatGPT.
- Connect and log in with your account info.
- Add the Hypermemory skill in your chat and start using Hypermemory.
Available Tools
| Tool | Description |
|---|---|
| hm_find_related | Traverse connected knowledge from a starting node |
| hm_forget | Delete a node with optional cascade |
| hm_get_overview | Graph stats, top nodes, and type distribution |
| hm_ingest | Decompose dense text into graph entities in one pass |
| hm_list_files | Query uploaded files by type or name |
| hm_list_orphans | List disconnected nodes for cleanup |
| hm_recall | Hybrid search across your graph (BM25, vector, edges) |
| hm_store | Create a memory node; background enrichment runs automatically |
| hm_timeline | Temporal recall by period or date range |
| hm_timeline_write | Write an explicit diary-style timeline entry |
| hm_update | Update a node; re-embeds when the description changes |
| hm_upload_file | Upload a file into memory (summary + graph node) |
Use Case
You capture objections and pricing notes in ChatGPT with hm_store. Monday you open Claude for a follow-up, Hypermemory returns the linked people and outcomes, not a blank thread. You chose the tool; the memory stayed put.