
Emem — Earth Memory Protocol
Give agents a signed, content-addressed memory layer for real-world places without managing your own geo database.
Overview
io.github.Vortx-AI/emem is a MCP server for the Build phase that connects agents to content-addressed, ed25519-signed Earth place memory over HTTP.
What is this MCP server?
- Content-addressed, ed25519-signed records—reads do not require API keys per project positioning
- Hosted remote at https://emem.dev/mcp (streamable-http) plus local Docker on port 5051
- Persistent sled cache + identity via EMEM_DATA volume mount
- Apache-2.0 licensed Earth memory protocol (emem v0.0.9)
- Optional EMEM_PUBLIC_URL for canonical self-referencing MCP responses
- Server manifest version 0.0.9
- Default container bind 0.0.0.0:5051 (EMEM_BIND)
- Apache-2.0 license; reads advertised as keyless
Community signal: 30 GitHub stars.
What problem does it solve?
You need trustworthy place-linked memory in your agent stack without building signing, addressing, and storage from scratch.
Who is it for?
Builders shipping location-aware agents, civic/geo tools, or signed audit trails tied to physical places.
Skip if: Simple CRUD apps that only need a generic document database with no geographic provenance.
What do I get? / Deliverables
After registration, your agent can read (and optionally host) signed place memory via streamable-http MCP with Docker or emem.dev remote.
- Agent access to Earth memory reads via MCP tools
- Local persistent emem node with sled cache on EMEM_DATA
- Integratable place identities for downstream app logic
Recommended MCP Servers
Journey fit
Place-memory protocols belong on the Build shelf as backend integrations when you wire location context into agents or apps. Integrations is where MCP servers connect external data protocols (HTTP remotes, Docker) into your product stack.
How it compares
Signed place-memory protocol MCP, not a generic SQL or Redis connector.
Common Questions / FAQ
Who is io.github.Vortx-AI/emem for?
Solo and small-team developers integrating verifiable geographic memory into MCP agents or backend services.
When should I use io.github.Vortx-AI/emem?
During Build when you are wiring integrations for maps, field data, or cross-session place context in an agent product.
How do I add io.github.Vortx-AI/emem to my agent?
Point your MCP client at https://emem.dev/mcp for the hosted streamable-http remote, or run the Docker image with EMEM_DATA mounted and connect to http://localhost:5051/mcp.