Skip to main content
Glama

MolMCP

Multi-plane MCP for the MolCrafts ecosystem.

Protocol: MCP 2026-07-28 via FastMCP 4.0.0b1 (+ MCP Python SDK v2). Handshake-era clients still work — FastMCP 4 negotiates per connection.

Optional science packages (molvis, molq, molexp, …): if not installed, that plane is omitted from catalogs and client configs (silent). Explicit molmcp serve <plane> still errors with an install hint. This is runtime behavior — not a test skip.

One product domain per MCP connection (separate process / server name). There is no mega-server under molmcp. Client default: all planes on. Turn planes off with --disable (and back on with --enable).

Plane

Command

Role

catalog

molmcp serve catalog

Bootstrap: list_planes, route(task)

molcrafts

molmcp serve molcrafts

Knowledge pages (packages → outline → open)

molvis

molmcp serve molvis

Live viewer session (open / exec / poll_events)

molq

molmcp serve molq

Job store + opt-in submit/cancel

molexp

molmcp serve molexp

Workspace layout + scaffold + data-directory adoption

Science APIs are never MCP tools. Discover them on the molcrafts plane, then call them from agent Python or inside molvis exec.

Client config (default: everything)

One standard mcpServers JSON, which every host reads — Claude Code and Cursor natively, Grok alongside its own config.toml.

molmcp client                            # all planes, to stdout
molmcp client --disable molq --disable molexp
molmcp client --disable molq --enable molq   # re-enable after a disable
molmcp client claude -o ~/.claude.json

An optional host (claude, cursor, grok) picks the default output path; the JSON itself is identical for all of them. A disabled plane is simply absent from the map. Tool ids look like molvis__open, not molmcp__molvis_open.

In Grok, ~/.grok/config.toml outranks the JSON sources. If an old molmcp entry lives there it still wins — grok inspect shows each server's origin.

Related MCP server: Octocode

Configuration

Settings live in ~/.molmcp/settings.json, edited through the CLI. There are no environment variables.

molmcp config list                                # resolved settings + layers
molmcp config set sources.molpy pkg:molpy         # index a package
molmcp config set indexWorkspace true --project   # index this repo too
molmcp config add excludes vendor
molmcp config remove sources.molpy

A project may carry .molmcp/settings.json (checked in) and .molmcp/settings.local.json (untracked); both layer over the user file. Writes go to the user file unless --project / --local is passed, because a plane server inherits its working directory from whichever client launched it.

What gets indexed. Auto-discovery finds installed MolCrafts distributions. The working directory is not a source unless indexWorkspace says so — it used to be, which meant an unconfigured install indexed whatever it happened to be started next to.

Key

Meaning

sources

Extra sources to index, name → spec (pkg:, local:, github:, path)

indexWorkspace

Index the working directory as well (default false)

knowledgeScope

Narrow which indexed sources the knowledge tools surface

excludes

Extra ignore globs for the file walk

cacheDir, maxCacheBytes, maxCacheAgeDays

Where the index lives and how big it may get

pythonEnv

Environment to auto-discover from (a venv root, python, or site-packages)

discoverInclude, discoverExclude

Force a distribution in or out of auto-discovery

molexp.workspace, molq.database

Provider-specific paths

molcrafts.json is no longer picked up from the working directory; pass --config PATH if you keep one.

CLI

uv run molmcp planes              # list planes
uv run molmcp client              # client config, all planes on
uv run molmcp config list         # resolved settings
uv run molmcp route "draw dopamine"
uv run molmcp serve catalog       # one plane per process
uv run molmcp serve molvis
uv run molmcp search "Conformer"  # offline index search
uv run molmcp index
uv run molmcp cache               # index size; --prune / --gc / --vacuum to reclaim

Install

uv sync --extra dev
uv run pytest -v

Design rules

  1. Multi-link on-demand — one process = one plane = one MCP server name.

  2. Bare tool names — the plane id is the server name, so a tool registers as open and the client shows molvis__open.

  3. No science tool mirror — no show_smiles / draw_dopamine; discovery + Python.

  4. Providers register via molmcp.providers entry points and are served with molmcp serve <name>.

  5. No environment switches — configuration is settings and CLI flags, so molmcp config list is the whole truth.

Documentation

Full manual: docs.molcrafts.org/molmcp (sources in docs/):

Local sources: docs/concepts/architecture.md, docs/guides/molvis-workbench.md.

A
license - permissive license
-
quality - not tested
B
maintenance

Maintenance

Maintainers
Response time
Release cycle
Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

  • A
    license
    -
    quality
    A
    maintenance
    Deep code indexing for AI agents. Search symbols, navigate call graphs, explore inheritance, track git history — all via MCP.
    54
    MIT
  • A
    license
    -
    quality
    A
    maintenance
    Semantic code indexer with GraphRAG knowledge graph. Index your codebase, search in natural language, and expose everything via MCP so AI agents understand architecture — not just files.
    445
    Apache 2.0
  • A
    license
    B
    quality
    A
    maintenance
    Local-first codebase intelligence engine providing AI coding agents with a typed MCP toolset for understanding and navigating code repositories.
    100
    51
    Apache 2.0
  • A
    license
    -
    quality
    A
    maintenance
    Provides code intelligence for AI coding agents by indexing repositories into a hybrid knowledge graph, enabling agents to query dependencies, impact, and context through 28 MCP tools.
    3
    Apache 2.0

View all related MCP servers

Related MCP Connectors

  • Shared, governed long-term memory for AI agents across tools and sessions via MCP and REST.

  • Persistent memory and cross-session learning for AI coding assistants (hosted remote MCP).

  • Cross-agent artifact workspace with provenance across Claude Code, Codex, Cursor, LangGraph.

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/MolCrafts/molmcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server