Neurohive
Coordinate memory and expertise across multiple agents so parallel coding agents do not contradict each other or lose who knows what.
Overview
neurohive is a MCP server for the Build phase that provides multi-agent memory intelligence with expertise tracking and conflict detection for coordinated AI agents.
What is this MCP server?
- Multi-agent memory intelligence with shared recall across parallel agent sessions
- Expertise tracking so agents know which specialist context owns a topic
- Conflict detection when stored memories or agent beliefs diverge
- npm stdio package (v1.0.2) for standard MCP client registration
- Designed for builder teams running more than one agent on the same codebase or product
- Server version 1.0.2
- Registry: npm package identifier neurohive
- Transport: stdio
Community signal: 1 GitHub stars.
What problem does it solve?
Running several agents in parallel makes it easy for contradictory memories and duplicate expertise to derail a solo builder’s codebase decisions.
Who is it for?
Indie builders orchestrating multiple MCP-connected agents on one product who need shared memory governance.
Skip if: Single-chat workflows with one agent and no cross-session memory coordination needs.
What do I get? / Deliverables
After wiring neurohive, agents can share tracked expertise and get conflict signals before acting on inconsistent stored context.
- stdio MCP tools for multi-agent memory read/write and intelligence queries
- Expertise metadata usable across agent sessions
- Conflict detection signals for inconsistent agent memory
Recommended MCP Servers
Journey fit
Multi-agent memory layers are implemented while you assemble agent stacks and orchestration, which sits in Build before you rely on them in production ops. neurohive targets agent-runtime intelligence (memory, expertise, conflicts), which maps to agent-tooling rather than generic backend APIs.
How it compares
Multi-agent memory orchestration MCP, not a single-user vector store or generic note app.
Common Questions / FAQ
Who is neurohive for?
It is for developers running multiple AI agents who need shared memory, expertise labels, and conflict awareness across sessions.
When should I use neurohive?
Use it when you split work across specialized agents and start seeing duplicated or conflicting long-term context.
How do I add neurohive to my agent?
Add the neurohive npm package (v1.0.2) as a stdio MCP server in your client configuration, install dependencies, then expose the server to agents that read and write hive memory per your workflow rules.