JSPM

  • ESM via JSPM
  • ES Module Entrypoint
  • Export Map
  • Keywords
  • License
  • Repository URL
  • TypeScript Types
  • README
  • Created
  • Published
  • Downloads 531
  • Score
    100M100P100Q90854F
  • License Apache-2.0

Continual learning MCP server for AI agents — topology-aware knowledge graph with kappa-routing, exposed as 5 loop-phase machines (retrieve/route/act/learn/consolidate), causal metadata, and neural embeddings. One of four MCP servers in the [&] three-protocol stack (box-and-box, graphonomous, os-prism, os-pulse).

Package Exports

    This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (graphonomous) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

    Readme

    graphonomous (npm wrapper)

    Continual learning memory loop for AI agents — 5 loop-phase MCP machines.

    One of four MCP servers in the [&] three-protocol stack:

    Package Role Install
    box-and-box [&] Protocol validator / composer npx -y box-and-box --db ~/.box-and-box/specs.db
    graphonomous Memory loop (this, 5 machines) npx -y graphonomous --db ~/.graphonomous/knowledge.db
    os-prism Diagnostic loop (6 machines) npx -y os-prism --db ~/.os-prism/benchmarks.db
    os-pulse PULSE manifest registry npx -y os-pulse --db ~/.os-pulse/manifests.db

    Graphonomous exposes the five memory-loop machines — retrieve, route, act, learn, consolidate — with κ-aware routing, causal metadata, and neural embeddings. When PRISM benchmarks Graphonomous, the loops nest: PRISM.interact → Graphonomous.retrieve → route → act → learn → consolidate → PRISM.observe.

    This package provides an npm-friendly launcher for the Graphonomous MCP server CLI.

    It installs (or reuses) a platform-specific graphonomous binary and exposes:

    • graphonomous ...
    • npx graphonomous ...

    The underlying server communicates over STDIO, so it works well with MCP-capable editors/clients (for example Claude Code, Zed, and Cursor).


    What this package does

    • Detects your OS/arch (darwin|linux + x64|arm64)
    • Downloads a matching release asset at install time
    • Installs the OTP release command path under vendor/<platform>-<arch>/graphonomous/bin/graphonomous when available
    • Creates/uses vendor/<platform>-<arch>/graphonomous as the launcher target for consistent execution
    • Runs the resolved Graphonomous command with all arguments passed through

    Requirements

    • Node.js >= 18
    • Supported platforms:
      • macOS: x64, arm64
      • Linux: x64, arm64

    Install

    Global install

    npm i -g graphonomous

    Then run:

    graphonomous --help

    One-off execution

    npx -y graphonomous --help

    Local project install

    npm i graphonomous
    npx graphonomous --help

    Run examples

    Start Graphonomous MCP server with a local DB path:

    graphonomous --db ~/.graphonomous/knowledge.db --embedder-backend fallback

    Safe laptop-oriented defaults:

    graphonomous \
      --db ~/.graphonomous/knowledge.db \
      --embedder-backend fallback \
      --log-level info

    Zed configuration example

    In Zed settings JSON:

    {
      "context_servers": {
        "graphonomous": {
          "command": "graphonomous",
          "args": ["--db", "~/.graphonomous/knowledge.db", "--embedder-backend", "fallback"],
          "env": {
            "GRAPHONOMOUS_EMBEDDING_MODEL": "sentence-transformers/all-MiniLM-L6-v2"
          }
        }
      }
    }

    If you prefer not to install globally:

    {
      "context_servers": {
        "graphonomous": {
          "command": "npx",
          "args": ["-y", "graphonomous", "--db", "~/.graphonomous/knowledge.db", "--embedder-backend", "fallback"],
          "env": {}
        }
      }
    }

    Claude Code configuration

    Add to your project's .mcp.json:

    {
      "mcpServers": {
        "graphonomous": {
          "command": "npx",
          "args": ["-y", "graphonomous", "--db", "./.graphonomous/knowledge.db", "--embedder-backend", "fallback"]
        }
      }
    }

    To run the full four-package stack side by side:

    {
      "mcpServers": {
        "box-and-box":  { "command": "npx", "args": ["-y", "box-and-box",  "--db", "~/.box-and-box/specs.db"] },
        "graphonomous": { "command": "npx", "args": ["-y", "graphonomous", "--db", "~/.graphonomous/knowledge.db", "--embedder-backend", "fallback"] },
        "os-prism":     { "command": "npx", "args": ["-y", "os-prism",     "--db", "~/.os-prism/benchmarks.db"] },
        "os-pulse":     { "command": "npx", "args": ["-y", "os-pulse",     "--db", "~/.os-pulse/manifests.db"] }
      }
    }

    Release asset override instructions

    By default, the installer resolves the GitHub release source from this package metadata (repository.url, then homepage, then bugs.url).

    Current default source in this package points to:

    • Owner: c-u-l8er
    • Repo: graphonomous

    So for version X.Y.Z, the default asset URL pattern is:

    • https://github.com/c-u-l8er/graphonomous/releases/download/vX.Y.Z/graphonomous-vX.Y.Z-<platform>-<arch>.tar.gz

    You can override this behavior with environment variables for custom repos/tags/asset hosting.

    Override GitHub owner/repo/tag

    GRAPHONOMOUS_GITHUB_OWNER=my-org \
    GRAPHONOMOUS_GITHUB_REPO=graphonomous \
    GRAPHONOMOUS_RELEASE_TAG=v0.1.1 \
    npm i graphonomous

    Override version used for asset naming

    GRAPHONOMOUS_VERSION=0.1.1 npm i graphonomous

    Use custom release base URL (bypass GitHub release URL construction)

    GRAPHONOMOUS_RELEASE_BASE_URL should point to a directory containing assets named like:

    graphonomous-v<version>-<platform>-<arch>.tar.gz

    Example:

    GRAPHONOMOUS_RELEASE_BASE_URL=https://downloads.example.com/graphonomous \
    GRAPHONOMOUS_VERSION=0.1.1 \
    npm i graphonomous

    Private release download token

    GRAPHONOMOUS_GITHUB_TOKEN=ghp_xxx npm i graphonomous

    (You can also use GITHUB_TOKEN.)

    Skip, force, and tune download behavior

    # Skip download entirely
    GRAPHONOMOUS_SKIP_DOWNLOAD=1 npm i graphonomous
    
    # Force re-download even if binary exists
    GRAPHONOMOUS_FORCE_DOWNLOAD=1 npm i graphonomous
    
    # Timeout and redirect controls
    GRAPHONOMOUS_DOWNLOAD_TIMEOUT_MS=120000 \
    GRAPHONOMOUS_DOWNLOAD_MAX_REDIRECTS=10 \
    npm i graphonomous

    Runtime command override

    You can bypass installed vendor binaries/release layout and point directly to a custom executable:

    GRAPHONOMOUS_BINARY_PATH=/absolute/path/to/graphonomous graphonomous --help

    Troubleshooting

    Binary not found after install

    Try reinstalling or rebuilding:

    npm rebuild graphonomous
    # or
    npm i graphonomous@latest

    Unsupported platform message

    Current prebuilt targets are Linux/macOS + x64/arm64.

    Permission issue on command path

    Reinstall the package, or manually set executable bit on unix-like systems:

    chmod +x node_modules/graphonomous/vendor/<target>/graphonomous
    chmod +x node_modules/graphonomous/vendor/<target>/graphonomous/bin/graphonomous

    PULSE manifest

    This package ships a PULSE loop manifest describing Graphonomous's 5-phase continual learning loop:

    • graphonomous.continual_learning.pulse.json (at the package root)

    Any PULSE-conforming benchmark (for example os-prism) can read this file directly to discover the loop's phases, signatures, substrates, and cross-loop token connections. To register it with a running os-pulse instance:

    npx -y os-pulse --db ~/.os-pulse/manifests.db
    # then, from another shell or an MCP client:
    #   pulse.register { "path": "$(npm root -g)/graphonomous/graphonomous.continual_learning.pulse.json" }

    Source of truth

    The npm package is a distribution wrapper around the Graphonomous Elixir CLI. Core implementation and release process live in the Graphonomous repository.