RecallMax Neural
Give multi-agent coding workflows a shared compressed memory so long swarm sessions stay coherent without blowing the context window.
Overview
RecallMax-Neural is an MCP server for the Build phase that compresses and recalls semantic context as a cognitive memory layer for LLM swarms.
What is this MCP server?
- Remote MCP over SSE at recallmax-mcp.vercel.app for swarm-oriented semantic context compression
- Cognitive memory layer designed for LLM swarms and multi-agent sessions
- Premium settlement via optional payment-signature header (EIP-3009 TransferWithAuthorization)
- Server schema version 1.0.0 under RecallMax-Neural title
- MCP integration you register in the client—not a packaged agent skill
- MCP server version 1.0.0
- 1 remote SSE endpoint (recallmax-mcp.vercel.app/api/mcp)
- 1 optional secret header: payment-signature (EIP-3009)
What problem does it solve?
Multi-agent coding sessions lose thread and burn tokens when every worker must reload the full conversation and artifacts from scratch.
Who is it for?
Indie builders running several agents on one product who need durable, queryable swarm memory without self-hosting a memory service first.
Skip if: Single-turn codegen with no swarm orchestration, or teams that only need static repo RAG instead of dynamic session memory.
What do I get? / Deliverables
After you register the RecallMax remote MCP endpoint, swarms can pull compressed shared memory so agents stay aligned across long builds without constant full-context replay.
- Registered remote RecallMax MCP connection in your agent
- Compressed semantic recall usable across swarm turns
- Premium-gated access when payment-signature is configured
Recommended MCP Servers
Journey fit
RecallMax is installed while you wire agent stacks and swarm orchestration—the Build phase is where cognitive memory layers attach to LLM tooling. Agent-tooling is the canonical shelf because the server is a semantic compression and memory substrate for LLM swarms, not a one-off integration or ship-time checker.
How it compares
Swarm cognitive-memory MCP integration, not a local brainstorming or planning agent skill.
Common Questions / FAQ
Who is RecallMax Neural for?
Solo and indie builders who orchestrate multiple LLM agents and need shared, compressed context across long coding sessions.
When should I use RecallMax Neural?
Use it during Build when you are wiring agent swarms, handoffs, or persistent memory so later turns do not depend on full chat re-feeds.
How do I add RecallMax Neural to my agent?
Add the remote SSE MCP server URL https://recallmax-mcp.vercel.app/api/mcp in Claude Code, Cursor, or another MCP client, and configure the payment-signature header if your tier requires premium settlement.