MCP server for IFC4.3 specification reference — search and retrieve IFC entity definitions, attributes, inheritance hierarchies, and PropertySets.
Unlike existing IFC-related MCP servers that operate on IFC model files (parse, extract, modify), this server provides access to the IFC specification itself as a structured reference. It enables AI to look up the correct definitions of entities, attributes, types, and constraints defined in the IFC4.3 standard (ISO 16739-1:2024).
Entity Search — find IFC entities by name or description keyword
Entity Definition — retrieve complete definitions with attributes, inheritance, WHERE rules, and documentation
Inheritance Tree — visualize ancestor chains and descendant hierarchies
PropertySet Lookup — get or search PropertySet definitions with full documentation
Dual Output — every tool supports both Markdown (human-readable) and JSON (structured) formats
Pagination — configurable limit/offset for large result sets
Related MCP server: rdf-mcp
Tool | Description |
| Search entities by name or description keyword |
| Get complete entity definition (attributes, inheritance, WHERE rules, docs) |
| Show inheritance hierarchy (ancestors, descendants, or both) |
| Get or search PropertySet definitions |
IFC4.3 specification reference: entity definitions, attributes, inheritance, WHERE rules, PropertySet definitions
Structured lookup of the IFC standard (ISO 16739-1:2024) — a vocabulary/dictionary for AI
Dual output (Markdown / JSON) suitable for both human reading and programmatic consumption
IFC file (.ifc) parsing or extraction — use IfcOpenShell (Python) or web-ifc (TypeScript/WebAssembly) instead
Geometry rendering or 3D visualization
Schema version migration (IFC2x3 / IFC4 → IFC4.3) — currently IFC4.3 only
Quantity information (Qto_*) — planned for a future release
If you need to operate on actual .ifc files at runtime, combine this MCP (for specification lookup) with a file-operation library (for data extraction).
npm install -g @shuji-bonji/ifc-core-mcpAdd the following to your claude_desktop_config.json:
{
"mcpServers": {
"ifc-core": {
"command": "npx",
"args": ["-y", "@shuji-bonji/ifc-core-mcp"]
}
}
}claude mcp add ifc-core -- npx -y @shuji-bonji/ifc-core-mcpgit clone https://github.com/shuji-bonji/ifc-core-mcp.git
cd ifc-core-mcp
npm install
npm run build
node dist/index.jsAfter npm install -g @shuji-bonji/ifc-core-mcp, the server can be started with:
ifc-core-mcpThe process communicates over stdio (MCP standard transport) and is intended to be launched by an MCP client (Claude Desktop, Claude Code, etc.) rather than invoked directly for interactive use.
Once installed, you can ask an MCP-enabled LLM (e.g., Claude) questions such as:
"What attributes does IfcSpace have in IFC4.3?"
"Show me the inheritance hierarchy from IfcBuildingElement down to leaf subtypes."
"Find all IFC entities related to HVAC."
"What PropertySets apply to IfcWall, and what properties do they contain?"
"What is the definition of IfcMapConversion, and which attributes store CRS information?"
The LLM will invoke the appropriate tool (ifc_search_entity, ifc_get_entity, ifc_get_inheritance, or ifc_get_propertyset) and return structured results.
The server covers the complete IFC4.3 schema:
Category | Count |
Entities | 876 |
Type Declarations | 132 |
Enumerations | 243 |
Select Types | 61 |
Functions | 48 |
Global Rules | 2 |
Entities are organized across 4 IFC layers:
Layer | Entities | Description |
Resource | 392 | Geometry, units, materials |
Core | 145 | Kernel, product extensions |
Shared | 111 | Walls, columns, beams, MEP |
Domain | 228 | Architecture, HVAC, rail, roads |
Build time (one-time, Python):
IFC.exp (EXPRESS schema) → ifc4x3-schema.json
IFC4.3.x-development Markdown → ifc4x3-descriptions-*.json
Runtime (TypeScript MCP server):
Pre-built JSON files → Map-based indexes → MCP toolsThe server loads three pre-built JSON data files at startup and builds in-memory indexes (Maps) for O(1) entity lookups by name. The EXPRESS schema provides structural data (types, attributes, inheritance, WHERE rules), while the Markdown documentation provides semantic data (definitions, usage, history).
EXPRESS schema (IFC.exp) — from buildingSMART/IFC4.3.x-output
Markdown documentation — from buildingSMART/IFC4.3.x-development
IFC specification content is licensed under CC BY-ND 4.0 by buildingSMART International
Node.js >= 22
Python 3 + IfcOpenShell (only for data regeneration)
npm run build # Compile TypeScript
npm run dev # Watch mode with tsx
npm test # Run all tests (unit + e2e)
npm run test:unit # Unit tests only
npm run test:e2e # E2E integration tests
npm run lint # ESLint check
npm run format # Prettier formatting
npm run prepare-data # Regenerate data from raw sources (requires Python)Runtime: TypeScript (strict), Node.js
MCP SDK: @modelcontextprotocol/sdk
Validation: Zod
Testing: Vitest (unit + e2e with actual MCP client)
Linting: ESLint + Prettier
CI/CD: GitHub Actions (lint → test → build → publish)
This server focuses on the IFC4.3 schema itself. Potential future directions under consideration:
QuantitySet (Qto_*) support — add quantity definitions alongside PropertySets
ifcJSON schema — complement the current EXPRESS-derived data
Cross-version lookup — map IFC2x3 / IFC4 → IFC4.3 entities
Sister MCP for IFC file operations — a separate project (planned, not yet released) for parsing and querying actual .ifc files, designed to pair with this reference server
Feedback and use-case reports are welcome via GitHub Issues.
w3c-mcp — MCP server for W3C/WHATWG/IETF web specifications
rfcxml-mcp — MCP server for IETF RFC documents (XML-based)
epsg-mcp — MCP server for EPSG coordinate reference systems (useful when combined with IfcMapConversion)
Get the complete definition of an IFC4.3 entity including attributes, inheritance, WHERE rules, and description.
Returns the entity's structural definition from the EXPRESS schema merged with the Markdown documentation.
Args:
name (string): IFC entity name (e.g. "IfcWall", "IfcBeam", "IfcProject")
include_inherited (boolean): Include attributes inherited from supertypes (default: true)
include_inverse (boolean): Include inverse relationship attributes (default: false)
include_description (boolean): Include Markdown description text (default: true)
response_format ('markdown' | 'json'): Output format (default: 'markdown')
Returns: Entity definition with attributes, types, inheritance chain, WHERE rules, and description.
Examples:
"IfcWall" → Wall entity with PredefinedType, supertype chain to IfcRoot
"IfcProject" → Project entity with global context attributes
| Name | Required | Description | Default |
|---|---|---|---|
| name | Yes | IFC entity name (e.g. 'IfcWall', 'IfcBeam', 'IfcProject') | |
| include_inverse | No | Include inverse attributes (default: false) | |
| response_format | No | Output format: 'markdown' for human-readable or 'json' for structured data | markdown |
| include_inherited | No | Include inherited attributes (default: true) | |
| include_description | No | Include Markdown description text (default: true) |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations declare readOnlyHint=true, idempotentHint=true, and destructiveHint=false, consistent with the description's 'Get' verb. The description adds detail on returned content (attributes, inheritance, WHERE rules, description) without contradicting annotations. No additional side effects or auth needs are mentioned, which is acceptable given the read-only nature.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is well-structured: a brief overview, clear bullet-style Args, Returns summary, and examples. Each sentence adds necessary information. No redundancy or fluff; ideal for an MCP tool.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the rich input schema and annotations, the description sufficiently explains the tool's purpose and behavior. It lacks an explicit output schema but summarizes return content. For a read-only query tool, this is adequate and complete.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Input schema covers all 5 parameters with block descriptions (100% coverage). The description repeats these in Args with minor additions (e.g., examples of entity names). This adds marginal value beyond the schema, meeting the baseline for high schema coverage.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool retrieves the complete definition of an IFC4.3 entity, listing included elements. It distinguishes implicitly from siblings like ifc_get_inheritance (which focuses on inheritance only) but does not explicitly differentiate. The purpose is clear and specific.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No explicit guidance on when to use this tool versus its siblings. The description provides examples but does not explain scenarios for choosing this over ifc_get_inheritance, ifc_get_propertyset, or ifc_search_entity. The agent must infer usage from the purpose alone.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get the inheritance hierarchy of an IFC4.3 entity.
Shows ancestor chain (supertypes up to IfcRoot) and/or descendant tree (subtypes).
Args:
name (string): IFC entity name (e.g. "IfcRoot", "IfcProduct", "IfcWall")
direction ('descendants' | 'ancestors' | 'both'): Which direction to show (default: 'both')
depth (number): Max depth for descendant tree, 1-10 (default: 3)
response_format ('markdown' | 'json'): Output format (default: 'markdown')
Returns: Inheritance hierarchy showing ancestors and/or descendants.
Examples:
"IfcRoot" with direction="descendants" → full class hierarchy tree
"IfcWall" with direction="ancestors" → IfcWall → IfcBuiltElement → ... → IfcRoot
"IfcProduct" with direction="both" → both ancestor chain and subtype tree
| Name | Required | Description | Default |
|---|---|---|---|
| name | Yes | IFC entity name (e.g. 'IfcRoot', 'IfcProduct', 'IfcWall') | |
| depth | No | Maximum depth for descendant tree (default: 3) | |
| direction | No | Direction: 'descendants' (subtypes), 'ancestors' (supertypes), or 'both' | both |
| response_format | No | Output format: 'markdown' for human-readable or 'json' for structured data | markdown |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, idempotentHint=true, and destructiveHint=false. The description complements this by explaining the tool's behavior (showing hierarchies, depth limits) without contradicting annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is well-structured with clear sections (main purpose, Args, Returns, Examples). Every sentence serves a purpose, though it is slightly lengthy. The use of examples aids understanding without adding unnecessary fluff.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's complexity (4 parameters, no output schema), the description provides decent coverage. However, the return format is vaguely described ('Inheritance hierarchy showing ancestors and/or descendants') and the examples only hint at the output structure. More explicit details about the output (e.g., format of markdown vs JSON) would improve completeness.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Input schema coverage is 100%, so baseline is 3. The description adds value by explaining parameter semantics through examples and clarifying default values and allowed ranges (e.g., depth 1-10, direction options). This goes beyond the schema's descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states that the tool retrieves the inheritance hierarchy (ancestors and/or descendants) of an IFC entity. It uses a specific verb ('Get') and resource ('IFC Inheritance Tree'), and the examples distinguish it from sibling tools like ifc_get_entity, which retrieves a single entity.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides clear context on when to use different parameters (direction, depth) and includes examples that illustrate typical use cases. However, it does not explicitly state when not to use this tool or mention alternatives among siblings, which would be helpful for an AI agent.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get or search IFC4.3 PropertySet definitions.
In 'get' mode, retrieves the full definition of a named PropertySet including individual property names, data types, and IFC type mappings. In 'search' mode, searches PropertySets by keyword.
Args:
name (string): PropertySet name (e.g. "Pset_WallCommon") or search keyword
mode ('get' | 'search'): Lookup mode (default: 'get')
limit (number): Max results for search mode (default: 20)
offset (number): Pagination offset for search mode (default: 0)
response_format ('markdown' | 'json'): Output format (default: 'markdown')
Returns: PropertySet definition with properties (name, dataType, ifcType, valueKind) and descriptions.
Examples:
name="Pset_WallCommon", mode="get" → Full PropertySet definition with 10 properties
name="Wall", mode="search" → All PropertySets related to walls
name="thermal", mode="search" → PropertySets with thermal properties
| Name | Required | Description | Default |
|---|---|---|---|
| mode | No | Mode: 'get' for exact name lookup, 'search' for keyword search | get |
| name | Yes | PropertySet name (e.g. 'Pset_WallCommon') or search keyword (e.g. 'Wall', 'thermal') | |
| limit | No | Maximum results to return | |
| offset | No | Number of results to skip for pagination | |
| response_format | No | Output format: 'markdown' for human-readable or 'json' for structured data | markdown |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already indicate read-only, idempotent, and non-destructive behavior. The description adds value by explaining the dual-mode behavior, pagination, and return structure, though it could detail more about edge cases.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is well-organized with clear sections (summary, args, returns, examples) and front-loaded with the purpose. It is slightly verbose by duplicating schema info, but remains readable and efficient.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the absence of an output schema, the description adequately explains return values and structure. It covers all parameters, modes, and pagination, making it suitable for effective tool use.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so the baseline is 3. The description repeats schema definitions and adds examples, which clarify usage but do not introduce new semantic meaning beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool retrieves or searches IFC PropertySet definitions, specifying the two modes and providing examples. It distinguishes itself from sibling tools like ifc_get_entity and ifc_search_entity by focusing on PropertySets.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description explains when to use 'get' vs 'search' with explicit examples, and the schema enforces mode choices. It does not explicitly contrast with siblings, but the tool's purpose is clear enough for appropriate selection.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Search IFC4.3 entities by name or description keyword.
Returns matching entities with their name, layer, schema, and short definition. Supports partial name matching and description text search.
Args:
query (string): Search keyword (e.g. "Wall", "beam", "spatial", "opening")
limit (number): Max results, 1-100 (default: 20)
offset (number): Pagination offset (default: 0)
response_format ('markdown' | 'json'): Output format (default: 'markdown')
Returns: List of matching entities with name, layer, schema, and short definition.
Examples:
"Wall" → IfcWall, IfcWallType, IfcWallStandardCase, ...
"spatial" → IfcSpatialElement, IfcSpatialStructureElement, ...
"beam" → IfcBeam, IfcBeamType, IfcBeamStandardCase, ...
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | Maximum results to return | |
| query | Yes | Search keyword to match against entity names or descriptions (e.g. 'Wall', 'beam', 'spatial') | |
| offset | No | Number of results to skip for pagination | |
| response_format | No | Output format: 'markdown' for human-readable or 'json' for structured data | markdown |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true, idempotentHint=true, destructiveHint=false. The description adds behavioral context such as partial name matching, description text search, and the return fields (name, layer, schema, short definition). It does not contradict annotations.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise and well-structured: a clear opening sentence, followed by a list of return fields, parameter descriptions, examples, and default values. Every sentence provides useful information without redundancy.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's complexity (4 parameters, no output schema), the description adequately explains the return format and pagination. It covers all necessary aspects for an agent to use the tool correctly, though adding explicit mention of the response format behavior would strengthen completeness.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Input schema has 100% coverage, so baseline is 3. The description adds value by providing example values for the query parameter (e.g., 'Wall', 'beam', 'spatial') and explaining pagination with limit and offset, which enhances understanding beyond the schema descriptions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states it searches IFC4.3 entities by name or description keyword, using a specific verb ('Search') and resource ('IFC Entities'). It distinguishes itself from sibling tools like ifc_get_entity (which likely retrieves a single entity) by focusing on search with partial matching and text search.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides clear context on when to use the tool (searching entities by keyword) and gives examples. It does not explicitly state when not to use or mention alternatives among siblings, but the use case is well-defined.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Each tool serves a distinct purpose: entity definition lookup, inheritance hierarchy, propertyset retrieval/search, and entity search. No functional overlap, clear boundaries.
All tools follow the pattern 'ifc_<verb>_<noun>', but verbs are not perfectly uniform ('get' for three, 'search' for one). Consistent prefix and noun naming, minor verb inconsistency.
Four tools is reasonable for a core IFC schema server. Covers essential lookups without being too sparse or bloated. Could include a few more (e.g., types, quantities) but current count is appropriate.
Covers key aspects of IFC schema: entity definitions, inheritance, propertysets, and search. Missing metadata like types list or quantity sets, but core read operations are well-covered.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
MCP server for Product Management
Hosted 3GPP MCP server for Rel-15–20 TS/TR search. Index stays current.
MCP server for AI access to Swagger by SmartBear.
MCP server for querying Forkast documentation
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/shuji-bonji/ifc-core-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
