Skip to main content
Glama

Brain Memory

Server Details

Persistent memory for AI agents: one brain shared across Claude, Cursor, Codex and more.

Ownership verified
Status
Healthy
Uptime
95.8% over the last 24 hours
OAuth
Works in Glama
Last Tested
Transport
Streamable HTTP · MCP 2025-11-25
URL
Repository
omelas-tech/brain
GitHub Stars
3
Server Listing
Brain Memory
ChatGPT
–
Claude
–

TDQS

A3.9/5.0

Scored across 8 tools

Disambiguation5/5

Each tool targets a distinct action: memorize (create), recall (read), forget (archive), pin/unpin (tier management), verify (quarantine resolution), status (health/listing), and use (active brain switch). The only near-overlap is forget vs. verify 'reject' (both archive), but the descriptions make the context-dependent distinction clear. No ambiguity in selection.

Naming Consistency5/5

All eight tools follow the same brain_verb snake_case pattern with verb-based naming (memorize, recall, forget, pin, unpin, verify, status, use). The prefix and casing are perfectly uniform, making the set predictable and easy to scan.

Tool Count5/5

Eight tools is well-scoped for a memory system, covering creation, retrieval, archival, tiering, quarantine review, status, and brain switching. Each tool earns its place; none appear redundant or trivial.

Completeness4/5

Core memory lifecycle is well covered: create, recall, archive, pin/unpin, verify, status. However, there is no update/edit tool to revise a memory's content, and although forget is described as recoverable (moved to _archived/), there is no restore/unarchive operation, leaving a minor dead end.

Publisher details

Operator
Omelas
Operator website
https://omelas.tech
Vendor relationship
First-party
Privacy policy
Not provided
Restrictions
Requires a Brain Cloud account (Google sign-in via OAuth). Free plan available; Pro is $1/month for more storage and brains.

Available Tools

8 tools
brain_forgetA
Destructive
Inspect

Archive a memory, by id, so it stops surfacing in recall (recoverable — moved to _archived/, not permanently deleted).

Inputs: id (required).

brain_memorizeB
Destructive
Inspect

Store a new memory containing exactly the text provided in content — a fact, decision, preference, or note. Only that text is stored. Returns the stored memory's id and path.

Inputs: content (required), tags, type, title, +5 more.

brain_pinA
DestructiveIdempotent
Inspect

Pin a memory, by id, to the always-present tier so it loads every session and never decays.

Inputs: id (required).

brain_recallB
Read-only
Inspect

Recall the user's most relevant stored memories for a query, ranked by the brain engine (TF-IDF relevance + decayed strength + spreading activation + context match).

Inputs: query (required), as_of, limit, project, +1 more.

brain_statusA
Read-only
Inspect

Health overview of the user's brain: memory count and last-updated time, and the account's brains in Brain Cloud with the one in use marked active.

Inputs: no named parameters.

brain_unpinA
DestructiveIdempotent
Inspect

Remove a memory, by id, from the always-present tier (returns it to normal recall + decay).

Inputs: id (required).

brain_useA
Idempotent
Inspect

Switch which of the account's Brain Cloud brains this connector reads and writes, by name or id (brain_status lists them). The choice is remembered for later sessions. Nothing is copied between brains.

Inputs: brain (required).

brain_verifyA
Destructive
Inspect

Review and resolve memories pending verification (writes from untrusted sources — tool output, external content, or instruction-shaped text — are quarantined until reviewed). action 'list' (default) is read-only; 'approve' clears the flag (origin and trust weighting stay); 'reject' archives the memory.

Inputs: ids, action.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    Try in Browser

    Glama MCP Gateway

    Add one secure layer between your agents and this server.