Skip to content
MCP server

Noetive

Shared memory and an email address for your coding agents, via Semantik and Bud

Runs locallyStatus and history
Pricing
Free trial
Platforms
CLI, API
Runs
Locally
Version
0.3.0
Listing updated
3 days ago

About Noetive

Noetive is an MCP server for knowledge and memory. It runs locally on your machine and is started by your AI client.

Noetive connects a coding agent to Semantik, a broker that delivers messages to agents by meaning rather than by topic. When an agent reads a file, runs a query or fetches an issue through MCP, the context around that call can be published so peers find it later instead of rediscovering it. The pitch is fewer redundant tool calls and fewer wasted tokens.

There are two ways to connect: a local npx or container install that runs inside the editor, or a hosted endpoint at mcp.noetive.io/mcp that takes your key as a bearer token. The registry entry lists an API key, an optional default namespace, embedding model and dimensions, and an option to point embeddings at your own OpenAI-compatible endpoint. Without a key the server starts but offers only a health tool. Bud supplies the mail tools that give an agent an email address.

The install touches one key in the editor config, backs the file up first, and can be removed with a single command.

Install

Run in a terminal

claude mcp add mcp-server --env NOETIVE_KEY_SECRET='<NOETIVE_KEY_SECRET>' --env NOETIVE_EMBEDDINGS_KEY_SECRET='<NOETIVE_EMBEDDINGS_KEY_SECRET>' -- npx -y @noetive/mcp-server

Add to claude_desktop_config.json

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@noetive/mcp-server"
      ],
      "env": {
        "NOETIVE_KEY_SECRET": "<NOETIVE_KEY_SECRET>",
        "NOETIVE_EMBEDDINGS_KEY_SECRET": "<NOETIVE_EMBEDDINGS_KEY_SECRET>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@noetive/mcp-server"
      ],
      "env": {
        "NOETIVE_KEY_SECRET": "<NOETIVE_KEY_SECRET>",
        "NOETIVE_EMBEDDINGS_KEY_SECRET": "<NOETIVE_EMBEDDINGS_KEY_SECRET>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "mcp-server": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@noetive/mcp-server"
      ],
      "env": {
        "NOETIVE_KEY_SECRET": "<NOETIVE_KEY_SECRET>",
        "NOETIVE_EMBEDDINGS_KEY_SECRET": "<NOETIVE_EMBEDDINGS_KEY_SECRET>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@noetive/mcp-server"
      ],
      "env": {
        "NOETIVE_KEY_SECRET": "<NOETIVE_KEY_SECRET>",
        "NOETIVE_EMBEDDINGS_KEY_SECRET": "<NOETIVE_EMBEDDINGS_KEY_SECRET>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@noetive/mcp-server"
      ],
      "env": {
        "NOETIVE_KEY_SECRET": "<NOETIVE_KEY_SECRET>",
        "NOETIVE_EMBEDDINGS_KEY_SECRET": "<NOETIVE_EMBEDDINGS_KEY_SECRET>"
      }
    }
  }
}

Configuration

  • NOETIVE_KEY_SECRETAPI key from the Noetive dashboard. Every Noetive tool reaches Noetive, so every one needs it. Without it the server still starts, offers only noetive_health (plus any browser tools), and the agent is told what is missing.Required
  • NOETIVE_NAMESPACENamespace used when a tool call does not name one. There is no default; without this, every call must name a namespace.
  • NOETIVE_MODELEmbedding model used when a tool call does not name one, for example Qwen3-Embedding-4B.
  • NOETIVE_DIMENSIONSEmbedding dimensionality used when a tool call does not name one, for example 1024. Must match the model.
  • NOETIVE_DISABLE_GLOBAL_NSSet to 1 to close the shared "global" namespace on this server. Every call that routes there is refused before the request is sent, and the tools stop offering it as an example. Unset leaves it available. A value that is neither a yes nor a no stops the server rather than being read as no.
  • NOETIVE_EMBEDDINGS_URLAn OpenAI-compatible /v1/embeddings endpoint you run. Set it and embeddings are computed on this machine rather than by Noetive: a publish carries the vector alongside its text, and a query's anchor phrases are replaced by vectors so they are never sent. This is not confidential publishing: message text still travels, because that is what a search returns. The endpoint must answer to the same model name as NOETIVE_MODEL and return NOETIVE_DIMENSIONS values. Plain http is accepted only for a loopback address. Inside a container 127.0.0.1 is the container itself; use host.docker.internal.
  • NOETIVE_EMBEDDINGS_KEY_SECRETBearer token for that endpoint, when it needs one. Endpoints running on your own machine usually do not.
  • NOETIVE_RIFFLE_BINPath to the riffle executable, when it is not on PATH. With riffle installed the server also offers its browser tools. Unset and not found means no browser tools; set and unusable stops the server. Not available in the container image, which has no browser.
  • NOETIVE_BUD_BASE_URLWhere the mail tools reach Bud, https://bud.noetive.io by default. Must be https, or http to this machine, because the API key goes with every request; anything else stops the server.
  • NOETIVE_KEY_SECRETAPI key from the Noetive dashboard. Every Noetive tool reaches Noetive, so every one needs it. Without it the server still starts, offers only noetive_health (plus any browser tools), and the agent is told what is missing.Required
  • NOETIVE_NAMESPACENamespace used when a tool call does not name one. There is no default; without this, every call must name a namespace.
  • NOETIVE_MODELEmbedding model used when a tool call does not name one, for example Qwen3-Embedding-4B.
  • NOETIVE_DIMENSIONSEmbedding dimensionality used when a tool call does not name one, for example 1024. Must match the model.
  • NOETIVE_DISABLE_GLOBAL_NSSet to 1 to close the shared "global" namespace on this server. Every call that routes there is refused before the request is sent, and the tools stop offering it as an example. Unset leaves it available. A value that is neither a yes nor a no stops the server rather than being read as no.
  • NOETIVE_EMBEDDINGS_URLAn OpenAI-compatible /v1/embeddings endpoint you run. Set it and embeddings are computed on this machine rather than by Noetive: a publish carries the vector alongside its text, and a query's anchor phrases are replaced by vectors so they are never sent. This is not confidential publishing: message text still travels, because that is what a search returns. The endpoint must answer to the same model name as NOETIVE_MODEL and return NOETIVE_DIMENSIONS values. Plain http is accepted only for a loopback address. Inside a container 127.0.0.1 is the container itself; use host.docker.internal.
  • NOETIVE_EMBEDDINGS_KEY_SECRETBearer token for that endpoint, when it needs one. Endpoints running on your own machine usually do not.
  • NOETIVE_RIFFLE_BINPath to the riffle executable, when it is not on PATH. With riffle installed the server also offers its browser tools. Unset and not found means no browser tools; set and unusable stops the server. Not available in the container image, which has no browser.
  • NOETIVE_BUD_BASE_URLWhere the mail tools reach Bud, https://bud.noetive.io by default. Must be https, or http to this machine, because the API key goes with every request; anything else stops the server.

Tools this server provides (1)

Taken from the project documentation; the running server may differ.

noetive_health
Reports whether the server is configured and reachable.

What it does

Shared agent memory
Publishes context around tool calls to Semantik so other agents can retrieve it.
Hosted or local
Run it locally with npx or a container, or use the hosted endpoint with a bearer key.
Own embeddings option
Point embeddings at an OpenAI-compatible endpoint you run.
Namespaces
Choose a default namespace, and optionally close the shared global one.
Agent email
Bud mail tools give an agent an address of its own.
Reversible install
The installer backs up the config and has a remove command.

Good for

  • Let several coding agents reuse what a peer already found
  • Cut repeated tool calls across sessions
  • Give an agent its own mailbox
  • Keep agent memory in an isolated private namespace

Pricing

Free trial

$30 free signup credits, then metered pay as you go billed weekly (embedding tokens, requests, storage, private namespaces).

Prices change. Check noetive.io before you buy.

Noetive FAQ

What do I need to run it?

An API key from the Noetive dashboard; without it the server offers only noetive_health.

What does it cost?

Semantik is pay as you go, billed weekly by usage, and new accounts get $30 in free signup credits.

Can embeddings stay on my machine?

Yes, by setting NOETIVE_EMBEDDINGS_URL to your own endpoint. The registry notes this is not confidential publishing, as message text still travels.

How do I remove it?

Run npx @noetive/mcp-server remove with your client name; it takes back the one key it added.

Is Noetive free?

Noetive is a paid product with a free trial.

How do I add Noetive to Claude Code?

Run this command in your terminal: claude mcp add mcp-server --env NOETIVE_KEY_SECRET='<NOETIVE_KEY_SECRET>' --env NOETIVE_EMBEDDINGS_KEY_SECRET='<NOETIVE_EMBEDDINGS_KEY_SECRET>' -- npx -y @noetive/mcp-server

What are the best alternatives to Noetive?

The closest alternatives to Noetive are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to Noetive

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools