llmbrain.ai · remote MCP server · beta
The memory your coding agents were missing.
One persistent brain across every project. Claude Code and Codex read it before starting, update it before ending — and never start from zero again.
- projects:
- all of them
- orientation:
- ~400 tokens
- deleted:
- nothing, ever
Requires an account and API key — create a free account →
The problem
Two repos. One product. Zero shared memory.
Your frontend agent has never met your backend. Connect both repos to the same llmbrain project and they share the docs that matter — api-contract, data model, milestones, last sessions.
“read api-contract.md — implement the client exactly as the backend shipped it”
Goodbye, /docs
Your /docs folder, promoted to the cloud.
Stop scattering markdown across repos where it rots unread. llmbrain holds the canonical docs — versioned, freshness-tracked, and readable by every agent in every repo.
Before / after
Same agent. Same you. Two very different mornings.
without llmbrain
- Starts siloed in one repo, blind to the rest
- No idea what was decided last week, or why
- You paste the same architecture summary. Again.
- Thousands of tokens burned re-establishing context
with llmbrain
- start_session("project") — status, open issues, where work stopped
- Reads the canonical docs, knows what's stale
- Picks up exactly where the last session stopped
- ~400 tokens, then straight to work
Semantic search
Ask your whole history a question.
Every doc, decision, and issue is indexed — keyword and vector search fused into one ranked answer, in milliseconds. Your agent finds the design doc from three weeks ago the same way you would: by meaning, not filename.
- Hybrid: full-text + vector, rank-fused
- Searches docs and issues together
- Results are addressable — llmbrain:// URIs agents can open
- Alexandruadmin
- Mira
- Dan
- Sofia
- shop3 sessions today
- robot-fw1 session today
- ops-toolingidle · 2d
Teams
One brain for the whole team.
Move a project into a team and every member’s agents read and write the same memory. What Mira’s agent learned this morning, Dan’s agent knows this afternoon — no handoff meeting, no stale wiki.
- Invite by email — members see every team project
- Admins manage members and move projects in and out
- Docs, issues, and sessions stay shared and fresh
A real session
One morning with llmbrain, start to finish.
- 09:02you
“Continue on shop-api — take the next slice from v0.3.”
- 09:02agent → llmbrain
Orients itself before touching code.
start_session("shop") → next slice: issue #44 last session: #43 closed — pagination on /orders - 09:03agent → llmbrain
Pulls the slice and its references.
get_issue(#44) → "webhook retries, exp backoff" ref: decision ADR-007 · doc api-contract §webhooks
- 09:05agent
Works the issue — code, tests, the actual job. llmbrain stays out of the way.
- 09:47agent → llmbrain
Closes the slice and freshens the shared contract doc.
update_issue(#44, done) · save_doc("api-contract") §webhooks retry semantics → fresh - 09:49memory written
Tomorrow's session — in this repo or the frontend — starts from here.
save_doc("status") → next slice queued: issue #45
Stop re-explaining your projects to your agents.
Create a free account and point your agents at the MCP endpoint — sign up with your email or straight through Google.
Create your accountnothing is ever deleted — your agents’ memory is yours