About Open Investment Model
Open Investment Model is a read-only MCP server for searching and referencing OpenIM, a model for institutional buy-side investment management. Results identify the OpenIM release and link to official sources; the server does not provide investment advice, execute transactions, or expose an implementation environment.
Key features
- 01Search model elements
- 02Retrieve an element by identifier
- 03Map architecture requirements to elements
- 04List released model exports
- 05Provide model identity and version
Use cases
- 01Find business and service domains
- 02Look up canonical entities and relationships
- 03Relate architecture requirements to model elements
- 04Review published schemas and ontologies
MCP support
The server provides five MCP tools for search, retrieval, requirement mapping, export listing, and OpenIM identity, plus the openim://identity and openim://catalog MCP resources.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
5Saved tool list · updated 2w ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| limit | integer | No | |
| query | string | Yes | Plain-language query or an OpenIM identifier such as SD-07.3 or E-04. |
| types | array | No | Optional model-element types to include. |
Details
From the project READMEOpenIM MCP server
Read-only, source-cited access to the Open Investment Model (OpenIM), the open reference model for institutional buy-side investment management.
The server searches and retrieves OpenIM Business Domains, Service Domains, Service Operations, canonical entities, relationships, glossary terms and generated model exports. Every response identifies the exact OpenIM release and links to the official versioned source. It does not provide investment advice, execute transactions or expose an implementation environment.
Run
Node.js 20 or later is required.
npx -y @openinvestmentmodel/openim-mcp
Generic MCP client configuration:
{
"mcpServers": {
"openim": {
"command": "npx",
"args": ["-y", "@openinvestmentmodel/openim-mcp"]
}
}
}
The server needs no API key and performs no network calls at runtime. Its catalogue is generated from the released OpenIM model when the package is built.
Tools
openim_search: search all model elements, optionally filtered by type.openim_get: retrieve one element by its OpenIM identifier.openim_map_requirement: find the model elements most relevant to a stated buy-side architecture requirement.openim_list_exports: list downloadable schemas, ontologies, graph files and other released artefacts.openim_get_identity: return the canonical OpenIM identity, version, scope and official sources.
The server also exposes openim://identity and openim://catalog as MCP resources.
Development
From the mcp/ directory in the OpenIM repository:
npm ci
npm run quality
The build derives dist/model-index.json from ../model, ../metadata/openim.json and ../exports. It bundles the reachable stdio runtime into dist/server.js. The package version in package.json is injected into that bundle and generates server.json.
The quality command runs the source production audit, unit and MCP tests, bundle inventory, packed consumer test, identity scan and reproducible pack check. The packed consumer calls all five tools through the installed tarball and checks the model version and official source in every response.
The package carries OpenIM's MIT licence in dist/LICENSE. Bundled third-party packages are listed with their licence texts in dist/THIRD_PARTY_NOTICES.txt. Linked legal comments remain beside the entry point in dist/server.js.LEGAL.txt.
Licence
MIT. OpenIM is a reference model, not a standard, investment product or production system.
Privacy Policy
This server runs locally on your machine. The model content ships inside the package; the server has no accounts, sends no telemetry, and operates no backend service. Full policy: https://openinvestmentmodel.org/privacy/
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.