Levirge Brain MCP server
Commit Oct 17, 2026Provides persistent shared memory for AI agents to store knowledge, skills, todos, and facilitate agent-to-agent handoffs.
Also known as levirge/brain, brain
- Stars
- 0
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- 26.10.17
- Last commit
- Oct 17, 2026
- License
- —
What is Levirge Brain MCP?
Levirge Brain provides persistent shared memory for AI agents to store knowledge, skills, todos, and facilitate agent-to-agent handoffs.
Summary generated from the project's README
About Levirge Brain MCP server
Levirge Brain provides a hosted knowledge base that allows AI agents to maintain persistent memory across different sessions. Instead of starting every interaction from scratch, agents use this server to store architecture decisions, runbooks, and technical lessons learned. It functions as a shared repository that prevents the need to re-explain your technology stack to every new session.
Developers can integrate this service into Claude Code, Codex CLI, or any MCP-compatible client like Cline. For remote setups, it uses a streamable HTTP transport via a specific URL. Users connect through OAuth sign-in on their first connection. The system includes specific tools such as search_knowledge for semantic retrieval, add_knowledge for capturing new information, and handoff capabilities to move work between different agents or roles.
This memory system focuses on grounding model responses through fact lineage and self-improving ranking. By using feedback to tune retrieval, the server adapts to how a team operates. It supports work sessions through commands that pull relevant context packs at the start of a task and persist new learnings when a session finishes.
What you can do with it
- 1Preventing repetitive context explanationsAsk the agent to remember specific architecture decisions so you do not have to repeat them in future sessions.
- 2Transferring tasks between agentsHand off ongoing work from one AI session or role to another using the shared inbox.
- 3Reducing model hallucinationsUse semantic search to ensure the agent cites actual documents rather than making up information.
- 4Managing task-specific contextInitialize a new work session with the correct context packs and save all findings once the task is complete.
Written from the project's README. Every use case is checked against a quote from it.
How to install Levirge Brain MCP
Formats verified 10 Nov 2026Add to CursorAdd to GooseInstall Levirge Brain MCP in Claude Code
$ claude mcp add --transport http brain https://brain.levirge.com/mcpInstall Levirge Brain MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"brain": {
"url": "https://brain.levirge.com/mcp"
}
}
}Install Levirge Brain MCP in VS Code
.vscode/mcp.json
{
"servers": {
"brain": {
"type": "http",
"url": "https://brain.levirge.com/mcp"
}
}
}Install Levirge Brain MCP in VS Code CLI
$ code --add-mcp '{"name":"brain","type":"http","url":"https://brain.levirge.com/mcp"}'Install Levirge Brain MCP in Codex
$ codex mcp add brain --url https://brain.levirge.com/mcpInstall Levirge Brain MCP in Codex config.toml
~/.codex/config.toml
[mcp_servers.brain] url = "https://brain.levirge.com/mcp"
Install Levirge Brain MCP in Gemini CLI
$ gemini mcp add -s user --transport http brain https://brain.levirge.com/mcpInstall Levirge Brain MCP in Gemini settings.json
~/.gemini/settings.json
{
"mcpServers": {
"brain": {
"url": "https://brain.levirge.com/mcp",
"type": "http"
}
}
}Automated checks
Ran Nov 15, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work · What to fix for more points
- Source repository linked and reachablelevirge/brain is public and reachablegithub.com/levirge/brain ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Not checked
- Open source license (OSI)No licenseFinding
- Commit or release in the last yearLast commit or release 2026-10-20github.com/levirge/brain ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNot checked
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeHTTPS endpoint answered 401 (sign-in required)Pass
README
View on GitHubBrain — persistent memory for your AI agents
Your agents forget everything between sessions. Brain doesn't.
Brain is a hosted knowledge base your AI agents read and write over MCP — architecture decisions, runbooks, hard-won gotchas, and agent-to-agent handoffs, searchable from Claude Code, Codex, and any MCP-compatible client. Stop re-explaining your stack to every new session.
Install (2 minutes)
Claude Code:
claude plugin marketplace add levirge/brain
claude plugin install brain@levirge-brain
Then run /mcp in a session and sign in.
Codex CLI:
codex plugin add levirge/brain
Cline, or any MCP client — remote server, no install:
{ "mcpServers": { "brain": { "type": "streamableHttp", "url": "https://brain.levirge.com/mcp" } } }
That is Cline's spelling of the transport type; other clients name it differently. OAuth sign-in on first connect. Details: llms-install.md.
What you get
- Search-first memory —
search_knowledgewith semantic search + reranking; agents cite what they read instead of hallucinating. - Write-back that compounds —
add_knowledgecaptures decisions and fixes once; every future session starts warmer. - Fact lineage — trace any answer back to the documents and atoms it came
from (
/brain:lineage). - Agent-to-agent handoffs — a shared inbox between agents and machines:
hand work from one session, role, or CLI to another (
/brain:handoff). - Work sessions & context packs —
/brain:startpulls the right context for a task;/brain:finishpersists what was learned. - Self-improving ranking —
feedback(helpful / dead-end verdicts) tunes retrieval to how your team actually works.
Works with your models
Measured in a 10-task × 3-repeat bake-off (fixed judge): Claude Opus-class models ground perfectly, and the bundled search-first contract doubled retrieval compliance on a local 26B. Details in plugins/brain/README.md.
Links
- App / sign in: https://brain.levirge.com
- Levirge: https://levirge.com
- Plugin reference: plugins/brain/README.md
Plugin versioning is semver, independent of the Brain server release.
Levirge Brain MCP questions
How do I add Levirge Brain MCP to Claude Code?
Run claude mcp add --transport http brain https://brain.levirge.com/mcp in your terminal. Add --scope user to make it available in every project.
How do I add Levirge Brain MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
How does the memory retrieval work?
It uses search-first memory with semantic search and reranking to ensure agents cite sources instead of hallucinating.
What is the authentication method?
The server uses OAuth sign-in upon the first connection.