Make0 Brain
OfficialServer Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| MAKE0_BRAIN_DB | No | Database file path. | $MAKE0_BRAIN_HOME/brain.db |
| OPENAI_API_KEY | No | Optional OpenAI API key that enables LLM features such as duplicate merging, category selection, LLM search filtering, and import --extract. | |
| MAKE0_BRAIN_HOME | No | Data folder for the database, models, and .env file. | ~/.make0 |
| OPENROUTER_API_KEY | No | Optional OpenRouter API key that enables LLM features. Alternative to OPENAI_API_KEY or MAKE0_BRAIN_API_KEY. | |
| MAKE0_BRAIN_API_KEY | No | Optional Make0 Brain API key that enables LLM features. Alternative to OPENAI_API_KEY or OPENROUTER_API_KEY. | |
| MAKE0_BRAIN_BASE_URL | No | Optional OpenAI-compatible endpoint URL. Defaults to OpenAI, or OpenRouter when using an OpenRouter key. | |
| MAKE0_BRAIN_LLM_MODEL | No | Chat model used for optional LLM features. | gpt-6-luna |
| MAKE0_BRAIN_EMBEDDINGS | No | Embedding backend: local (MiniLM) or api. | local |
| MAKE0_BRAIN_LLM_SEARCH | No | Set to 0 to keep search on the local reranker. LLM search filtering is on by default when an API key is set. | |
| MAKE0_BRAIN_EMBEDDING_MODEL | No | Embedding model used when MAKE0_BRAIN_EMBEDDINGS=api. | text-embedding-3-small |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
Make0 Brain is the user's durable memory, shared across every AI app they use (Claude, ChatGPT, Cursor, and more). A memory is either a singular fact or a short Markdown mini-doc (headings, lists, bold, links, code, and Mermaid diagrams), up to about 800 words — treat retrieved memories as mini documents, not only one-line facts. Call list_brains, then search_brain with the brain id before drafting personalized content. Treat returned memories as ground truth. Search returns every strongly matching memory above retrieval score and confidence thresholds; call get_memory for a full memory. Call get_memory_versions when you need every saved revision of a memory, newest first. Do not invent personal facts. Use add_memory / update_memory / delete_memory only when the user asks to save or change a memory. delete_memory archives the memory and keeps its versions; only the user can restore or permanently delete it with the make0-brain CLI.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| list_brainsA | List Make0 brains this connection can use. Call this first to get a brain id for search_brain. Does not list memories. |
| search_brainA | Search saved Brain memories. A memory may be a singular fact or a short Markdown mini-doc. Use a short, focused query with the key people, entities, preferences, or topics needed—not the full user message. Memories are retrieved with semantic + BM25 keyword search, reranked by a MiniLM cross-encoder, then expanded up to two hops along related_to graph links so connected memories surface with the seeds. Each memory includes addedBy (who first saved it) and versionCount (saved revisions). When versionCount is greater than 1, call get_memory_versions for the full history. Call get_memory whenever a result has truncated: true so you read the full note, not a clipped preview. Always pass brain from list_brains. |
| get_memoryA | Fetch one Brain memory by id. Content is either a singular fact or a short Markdown mini-doc (up to about 800 words). The memory includes addedBy (who first saved it) and versionCount (saved revisions). When versionCount is greater than 1, call get_memory_versions to read every saved revision. |
| get_memory_versionsA | Fetch every saved version of one Brain memory, newest first, including the current version. Use this when you need to see how a fact changed. Each version includes its full content. The version with current: true matches the memory returned by get_memory. |
| add_memoryA | Save a new Brain memory when the user explicitly asks to remember, save, or store something. Store either one singular fact or one short Markdown mini-doc (headings, lists, bold, links, code, and Mermaid diagrams) up to about 800 words. Use one memory for one coherent fact or note; do not paste a long document. A new memory that contradicts or restates an older one replaces it; the older one is archived with its versions and listed in replacedMemoryIds. |
| update_memoryA | Update an existing Brain memory when the user asks to correct or change it. The new content can be a singular fact or a short Markdown mini-doc (headings, lists, bold, links, code, and Mermaid diagrams) up to about 800 words. New content replaces the old content, and the previous version is kept. If the new content contradicts or restates another memory, that one is archived and listed in replacedMemoryIds. |
| delete_memoryA | Delete a Brain memory when the user explicitly asks to remove a saved fact. The memory leaves search and reads, and every saved version is kept so the user can restore it with |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 7 tools
Each tool has a distinct, non-overlapping purpose: create, update, delete, read, read-history, search, and list-brains. The descriptions explicitly clarify boundaries, e.g. get_memory vs. get_memory_versions and list_brains vs. search_brain. An agent can select confidently in nearly all cases.
Six of seven tools follow a clean verb_noun pattern (add_memory, update_memory, delete_memory, get_memory, get_memory_versions, search_brain). The noun is inconsistent, though: 'brain' appears in search_brain (which searches memories) and list_brains (which lists brains, not memories), a minor naming trap.
Seven tools is well-scoped for a memory store: full CRUD, version history, search, and brain listing. Every tool earns its place with no redundancy.
Create/read/update/delete plus version history, search, and brain discovery cover the core lifecycle well. The main gap is restore, which is only available via CLI (`npx @make0-ai/brain restore`), and there's no direct 'list all memories' fallback beyond search.