Skip to content

The TerraMCP directory

Find your next source.

Ranked by evidence. Filter by discipline, region and format to find sources for your project.

Reset filters

109 sources

Sponsored placements first · Organic quality score ↓

mcp
83/100Currency A

gods-eye-view ↗

A spy satellite simulator in your browser, except the data is real. Live open source spatial intelligence on a photorealistic 3D globe.

What's in it

Data offered docs-verified
live aircraft · ships · satellites · earthquakes · traffic · public cameras
Coverage llm-inferred
Global (globe-wide); public camera feeds limited to Austin, London, California, Finland and Delaware
Update frequency docs-verified
Live or regularly refreshed; per-feed intervals not documented
Access docs-verified
Free and open source (licence: Other, per metadata); runs locally in a browser. Optional keys: Cesium ion token (free, eligible personal non-commercial use), Google Maps key (paid, metered), OpenAI key (voice control only)

Useful for

Use cases llm-inferred
Desk-based site context overview (terrain, imagery, nearby live signals) before a site visit · Weather context review for an event window using forecast wind and observed radar replay · Visual briefing material for client and stakeholder presentations using scene capture
Audience llm-inferred
GIS analysts · flood consultants (visualisation only) · planners · civils/BIM engineers (visualisation only)

Limitations

Constraints llm-inferred
Traffic is simulated along real roads from aggregate location data, not measured · CCTV camera poses and rocket launch trajectories are coarse estimates · Public camera feeds only over listed regions · Google Maps and Cesium ion features are subject to provider terms and quotas · Voice control requires an OpenAI key · Bundled Nepal flood data is non-commercial only · Licence listed as Other, so reuse terms need checking before commercial use · Older versions no longer work with OpenStreetMap data sources and need updating

Globalhealthy

mcp
80/100Currency A

mcp-server-for-revit-python ↗

mcp-server-for-revit-python

What's in it

Data offered docs-verified
Revit model information (document title, model details) · Levels with elevations · View export as image · Family type listing and family placement · Current view info and visible elements · Parameter-based colour overrides (colour splash) and category parameter listing · Revit installation discovery, launch, open/close/save, central sync · IronPython execution inside the running Revit session
Update frequency llm-inferred
On demand: each tool call queries the live Revit session
Formats docs-verified
MCP stdio · MCP streamable HTTP · MCP SSE
Access docs-verified
Local install alongside Autodesk Revit and pyRevit Routes; no API key. Requires Windows-user pairing via MCP > Bridge > Connect. Licence MIT (catalogue metadata, not docs).

Useful for

Use cases llm-inferred
Model QA: list levels and elevations to check floor levels and invert assumptions against drainage or civils design · Parameter completeness check: colour-splash elements by parameter value to spot missing data before issue · Coordination drawings: export named Revit views as images for design review or client markup
Audience llm-inferred
BIM/civils engineers · BIM coordinators · Drainage and civils engineers using Revit · Python developers extending Revit automation

Limitations

Constraints docs-verified
Not yet Autodesk Marketplace certified · Requires desktop Revit running with pyRevit Routes active · Bridge accepts loopback hosts only; remote VM/LAN not supported · Docs status table marks 10 of the listed tools as Pending (e.g. create_line_based_element, delete_elements, modify_element) · Docs say the review candidate has 20 tools, but catalogue metadata lists 30 · pyRevit Routes API is in draft form and subject to change · execute_revit_code runs arbitrary Python; model-changing requests need in-Revit review

Globalhealthy

mcp
78/100Currency A

rvt-mcp ↗

BIMwright — AI-native Revit MCP server (Revit 2022–2027)

What's in it

Data offered docs-verified
Model query: view, selection, filters, stats, parameters, relationships, worksets, groups/assemblies · Element creation: grids, levels, rooms, line/point/surface-based elements, groups · Views and sheets: view creation, sheet layout helpers, image capture, crop/scale · Annotation, schedules, materials, MEP, sheets, rooms, structural, and export helpers (per toolset table) · C# escape hatch via revit_send_code_to_revit when no typed tool fits
Formats docs-verified
MCP stdio
Access llm-inferred
Free/open source, Apache-2.0 licence (per ENTRY). Runs locally on Windows alongside an installed Autodesk Revit 2022–2027. No credentials documented in README. Data stays on the machine.

Useful for

Use cases llm-inferred
Pre-issue model check before sheet issue (revit_pre_issue_check) · Model audit of a Revit project for QA (revit_model_audit) · Drawing and sheet layout automation for issue sets (revit_drawing_layout) · Stair modelling and checks (revit_stairs)
Audience llm-inferred
BIM coordinators · BIM/civils engineers · Revit modellers

Limitations

Constraints llm-inferred
Windows only: the installer and PowerShell setup target a Windows machine · Revit must be open with a model, and MCP must be started from the ribbon each session · Revit versions limited to 2022–2027 · Early implementation, researched and implemented since March 2026 · Family Editor authoring is out of scope · Local only: no remote access; connects to Revit via localhost TCP (≤2024) or a named pipe (≥2025)

Globalhealthy

mcp
77/100Currency A

RevitMCP ↗

model context protocol for Autodesk Revit

What's in it

Data offered llm-inferred
live project metadata (name, document path, Revit version) · active view info and bounded element snapshots · element parameters, relationships and locations · schedules and schedule rows · view and sheet listings · model statistics (category, family/type, level, room counts) · view and element images for vision inspection
Update frequency llm-inferred
On demand: live query of the open Revit session
Formats docs-verified
MCP stdio
Access llm-inferred
Free, MIT licence (per entry metadata); requires local install of Autodesk Revit and pyRevit with a Revit project open; no credentials documented

Useful for

Use cases llm-inferred
Pre-issue QA: audit view and sheet naming conventions across a drainage or civils model · Schedule checks: pull schedule rows and compare schedules for quantity reconciliation · Parameter QA: read element properties and flag missing values before exporting data · Visual review: export a view image of a drainage run or structure for markup in a design review
Audience llm-inferred
BIM/civils engineers · BIM managers and coordinators · drainage engineers · BIM/CAD technicians

Limitations

Constraints llm-inferred
Needs a live Revit session with a project open; no offline or file-based access · Requires pyRevit and the Revit Routes server enabled · Web UI binds to 127.0.0.1 by default; LAN sharing is for trusted networks only · Claude Desktop local MCP runs on one machine only · Schedule formula text is unsupported (public Revit 2024 API only) · Install instructions are Windows-specific · Tool argument schemas are not documented in DOCS

Globalhealthy

mcp
77/100Currency A

aec-model-bridge ↗

Open-source Revit MCP server: let Claude and other AI assistants read and edit BIM models, with human approval on every change.

What's in it

Data offered docs-verified
Revit model read and edit (elements, parameters, views, sheets, schedules) · IFC file structure, properties and validation (no Revit needed) · Rhino and Grasshopper geometry, layers and materials · Speckle projects, models, versions, send and receive
Formats docs-verified
MCP stdio
Access docs-verified
Open source under GPL-3.0 (per entry metadata), with a commercial licence also referenced in the docs. Live Revit use needs Windows and a licensed Revit 2024 to 2027; mock mode needs neither.

Useful for

Use cases llm-inferred
Model QA before design coordination: list Revit warnings and the affected element IDs · Parameter corrections on drainage or structural elements, applied only after human-approved plans · Read-only IFC storey and property checks on consultant models without opening Revit
Audience llm-inferred
BIM coordinators · BIM managers · BIM/civils engineers

Limitations

Constraints docs-verified
Pre-release: nothing has been verified in a live Revit session yet · Describes the dev branch; the latest GitHub release may not include all features · Live Revit work limited to Windows and Revit 2024 to 2027 · Approving a plan changes the open model; test on a copy first · Snapshot QA/QC and report modules need snapshot_id; model_bloat lacks the guard · Advanced tools (revit_execute_python, revit_invoke_method) have the full permissions of the Revit process · Navisworks support is in progress · No UK-specific standards, datasets or civils tooling described in the docs

Globalhealthy

mcp
75/100Currency A

arcgis-mcp-bridge ↗

Secure, local-first MCP server exposing ArcGIS Pro's ArcPy engine over stdio JSON-RPC.

What's in it

Data offered docs-verified
Map algebra · zonal stats · DEM slope/aspect/hillshade · hydrology · kernel density · Gi* hot spots · Moran's I · service areas and routing · OD cost matrix · closest facility · PDF/PNG map plots · FC/GDB lifecycle · overlays, dissolve/merge · CRS define/project for vector and raster
Formats docs-verified
ArcGIS geodatabase (.gdb) feature classes · ArcGIS Pro .aprx · Excel · GeoJSON · CSV · PDF · PNG · MCP stdio (JSON-RPC 2.0)
Access docs-verified
Requires Windows and a licensed ArcGIS Pro installation (Esri licence), plus Python 3.11+. Apache-2.0 licence. No credentials documented in README.

Useful for

Use cases llm-inferred
Slope and aspect from a site DEM for drainage or earthworks setting-out (raster_operations) · Overlay and dissolve constraint layers such as flood zones and site boundaries in a project GDB (geometry_analysis) · Hot-spot or kernel density analysis of flood or incident records (spatial_statistics) · Export PDF/PNG map plates for planning or flood risk assessment submissions (export_layout)
Audience llm-inferred
GIS analysts · drainage engineers · hydrologists · civils/BIM engineers · planners

Limitations

Constraints docs-verified
Windows only; needs a licensed ArcGIS Pro installation and its Python environment set as ARCPY_PYTHON_PATH · Automated tests use mocked arcpy; the 100 geoprocessing tools are not validated across scenarios on a licensed runtime · Ten state-mutating tools (e.g. delete_dataset, repair_geometry) refuse to run without confirm: true · Some tools need Esri Spatial or Network extension licences · Sketch-to-GIS demo is marked 'coming soon'

Globalhealthy

skill
75/100Currency A

archsight-aios ↗

面向建筑行业知识工作从业者与 AI 研发团队的 Skills、Workflow 与多 Agent 工具包 / Building-industry AI agent skills for BIM, IFC, RAG, GraphRAG, project evidence work, code review, and runtime governance.

What's in it

Data offered docs-verified
Knowledge Pack: compiled engineering knowledge from norm excerpts, enterprise standards, project materials, historical review criteria and review records · Local Reference Runtime norm lookup (knowledge.norm_lookup) returning citations, versions, applicability and conflict status · Markdown writing workbench for tender, scheme and disclosure documents
Formats docs-verified
Markdown
Access docs-verified
Open source; licence Apache-2.0 (per ENTRY metadata); no credentials documented in README

Useful for

Use cases llm-inferred
Tender technical-bid compliance matrix, missing-document list and bid-risk flags from tender files (aios-tender-audit) · Special construction scheme review: hazard list, disclosure points and calculation-sheet gaps (aios-scheme-audit) · Building a norm-excerpt Knowledge Pack with citations for internal lookup (aios-knowledge)
Audience llm-inferred
construction tender and technical-bid staff · project document controllers · technical leads for construction schemes · AI/software teams building construction tools

Limitations

Constraints docs-verified
Knowledge Pack is not a production-grade norms database · Does not replace legal, chief engineer, expert, review or registered-professional sign-off · Docs describe Chinese-language construction workflows; no UK standards, water or civils coverage is documented · Not intended for non-construction tasks; outputs need human review

Globalhealthy

mcp
75/100Currency A

best-cad-mcp ↗

Comprehensive AutoCAD MCP server for Windows. It exposes high-level tools for drawing, editing, querying, annotation, layout, plotting, metadata, 3D solids, and CAD workflow guidance through the Model Context Protocol.

What's in it

Data offered docs-verified
2D drafting entities (lines, polylines, curves, circles, regions, hatches, text, dimensions, leaders, tables, layers, blocks, attributes) · editing operations (move, copy, rotate, scale, mirror, offset, trim, extend, fillet, chamfer, arrays) · SQLite scan and CAD-IR drawing understanding · semantic objects, constraints and dimension binding · 3D solids and operations · PDF/DXF/DWF/image export
Formats docs-verified
MCP stdio · DWG (input) · DXF · PDF · DWF · image
Access docs-verified
Free, MIT-licensed open source (licence from ENTRY metadata); requires AutoCAD installed and licensed on Windows, running as the same Windows user as the MCP client

Useful for

Use cases llm-inferred
Preflight-checked inspection and summary of existing DWG drawings before civils or drainage design review · Guarded, dry-run-verified layer, annotation and dimension clean-up of CAD deliverables before issue · Generating PDF plot and PNG visual evidence of drainage or civils layouts for internal QA
Audience llm-inferred
civils engineers · drainage engineers · BIM and CAD technicians · GIS and CAD analysts

Limitations

Constraints docs-verified
Beta software, intended for controlled local use with operator review rather than unattended changes to production drawings · Windows only; requires AutoCAD running as the same Windows user · Not claimed to comply with formal ISO, GB or ASME standards · Requires a local MCP client; no hosted or remote operation

Globalhealthy

mcp
74/100Currency A

local-austin-mcp ↗

Free, open-source MCP server (Apache-2.0) giving Claude plain-English access to official Austin & Travis County data — property records, permits, flood zones, city code text, schools, plus active MLS listings. Built by Neuhaus Realty Group.

What's in it

Data offered docs-verified
active MLS for-sale listings (active and under-contract only) · neighbourhood lookup · Travis, Williamson and Hays CAD property records · county tax bills, exemptions and delinquencies · MUD/PID special-district overlay · FEMA NFHL flood zones · City of Austin permits · code-compliance cases · zoning, lot dimensions and plats · municipal code text (Austin, plus Leander, Round Rock, Dripping Springs) · tree permits · 311 requests · APD crime, arrests, use-of-force and dispatch data · AFD dispatches · council voting records · city budget · council district, school district, ESD and voter precinct lookup · TEA school ratings and AISD attendance zones · library branches · parks · fire stations · restaurant inspection scores · roadway work zones · Animal Center intakes and outcomes · TxDOT highway projects · trash/recycling collection days · CapMetro real-time bus and rail · NWS alerts · Lake Travis and Highland Lakes level and 30-day trend · drive time and distance between addresses
Coverage docs-verified
Austin, Texas metro: City of Austin plus Travis, Williamson and Hays counties
Update frequency docs-verified
Live, per request: each tool queries its upstream source at call time; no fixed update cadence is documented
Formats docs-verified
MCP stdio
Access docs-verified
Free and open source under Apache-2.0; no API key or login required. Optional AUSTIN_SODA_APP_TOKEN (Socrata app token) raises the rate limit on data.austintexas.gov and datahub.austintexas.gov.

Useful for

Use cases llm-inferred
Flood zone, permit and code-case screening for a site due-diligence note · Zoning, lot dimensions and municipal code text lookup for a pre-application brief · Tree-permit and roadway work-zone check before site clearance or access planning · Special-district (MUD/PID) and appraisal context for a development or drainage-authority brief
Audience llm-inferred
flood consultants · planners · civils engineers · GIS analysts

Limitations

Constraints llm-inferred
Austin, Texas and Travis, Williamson and Hays counties only; no UK data, so relevance to a UK consultancy is low, although the catalogue lists regions as Global · Early-stage project run by a single maintainer, so expect gaps and possible breaking changes · MLS data covers active and under-contract listings only, with no sold comps · Listings, neighbourhood and blog content depend on neuhausre.com and the Neuhaus VOW API · Socrata rate limits apply unless an app token is set · Commute tool uses demo routing servers with no real-time traffic · Catalogue auth field says an API key is required, but docs say no key is needed · Catalogue tool list does not match docs: docs state 43 tools and list austin_utility_providers, austin_relocation, austin_local_voices and austin_health, which are absent from the metadata, while the metadata lists health_check

Some details auto-extracted · unverified

Globalhealthy

mcp
74/100Currency A

naviscoord-mcp ↗

Open-source MCP server for Autodesk Navisworks Manage 2024-2026: AI-assisted BIM clash detection, root-cause analysis, coordination plans and verified model workflows.

What's in it

Data offered docs-verified
ranked clash issues with deterministic scores and reasons · spatially bounded clash groups and duplicate collapse with traceable original clashes · root-cause hypotheses with measured evidence and affected occurrences · persistent issue identities, saved snapshots and delivery comparisons · work packages, trade ownership, narrative reports and PDF images · authoring element IDs and structured handoff files for downstream tools
Formats docs-verified
PDF images · structured handoff files · tables for downstream tools
Access docs-verified
Open source under MIT licence. The packaged local server needs no Python install, API key or build tools. Live coordination requires Windows and a licensed Autodesk Navisworks Manage 2024, 2025 or 2026; the AI client and Autodesk licence are separate products.

Useful for

Use cases llm-inferred
Prioritise open Navisworks clashes before a coordination meeting · Collapse repeated clashes to root-cause groups for trade discussions · Compare issue status between successive federated model deliveries · Prepare coordination work packages with trade ownership and PDF report images
Audience llm-inferred
BIM coordinators · BIM/civils engineers · design coordination managers

Limitations

Constraints docs-verified
Live coordination requires Windows and a licensed Autodesk Navisworks Manage 2024, 2025 or 2026 · Engineering approval and edits remain with the project team · Movement alternatives use element bounds and must be checked against surrounding geometry and design constraints · Revit and Power BI handoff relies on separate Horizun MCP servers

Globalhealthy

mcp
74/100Currency A

nepa-mcp ↗

PermitAI NEPA MCP Toolkit: MCP servers for federal environmental data, regulatory research, and geospatial screening to support permitting workflows.

What's in it

Data offered docs-verified
National Flood Hazard Layer flood zones, levees, and water areas · ESA-listed salmon and steelhead ranges by HUC-12 watershed · SSURGO soil map units and farmland classifications · ACS 5-Year socioeconomic indicators · EPA Air Quality System monitoring data · IPaC species, critical habitat, migratory birds, wetlands
Coverage llm-inferred
United States only (federal and some nonfederal sources); no UK coverage stated
Formats docs-verified
GeoJSON · HTML
Access docs-verified
No credentials needed for 20 servers; census requires CENSUS_API_KEY; epa_aqs requires EPA_AQS_EMAIL and EPA_AQS_API_KEY; BSD-3-Clause licence

Useful for

Use cases llm-inferred
Flood-zone constraint screening for a project area using fema_nfhl · Soil and farmland siting indicators for drainage or SuDS options using nrcs_soils · Species and critical-habitat screening for early ecological constraints using ipac · Composite constraints map of a project area for early site appraisal using map_composer
Audience llm-inferred
GIS analysts · flood consultants · ecologists · planners · civils and BIM engineers

Limitations

Constraints docs-verified
Screening and research aid only; does not make legal or agency determinations or replace agency or Tribal consultation · Does not guarantee that upstream datasets are complete or current · map_composer queries public GIS services at request time, so results depend on those services being available · census and epa_aqs return no data until their credentials are set

Example tool calls docs-verified

Retrieve IPaC species and critical-habitat records for early ecological constraints

{
  "tool": "ipac",
  "arguments": {
    "roi": "<project-area geometry; parameter name not stated in DOCS>"
  }
}

Some details auto-extracted · unverified

United Stateshealthy

mcp
73/100Currency A

sailing-mcp ↗

Remote MCP server for sailing navigation — wind/wave forecasts, tides, currents, marinas, bridge schedules and live AIS. Stateless, multi-tenant (users bring their own API keys).

What's in it

Data offered docs-verified
wind forecasts · wave forecasts · tides (HW/LW, observed, forecast, surge) · tidal currents · marina, harbour and anchorage POIs · scheduled bridge openings · Notices to Skippers · fairway depths and bridge air-draft clearance · locks · METAR/TAF aviation weather · sunrise, sunset, twilight and moon phase · live AIS vessel snapshot · global tide predictions (key required) · multi-source tidal/ocean currents (key required)
Coverage docs-verified
Dutch / North Sea waters (strong focus); some tools are global
Formats docs-verified
MCP streamable HTTP
Access docs-verified
Public-data tools need no key. Windy, WorldTides, Stormglass and AISStream need the user's own API key passed in the URL; all except Windy have a free tier. Server licence MIT (per ENTRY metadata, not DOCS).

Useful for

Use cases llm-inferred
Tide and surge levels at Dutch coastal sites for marine works or outfall boundary checks (tides_nl) · Wind and wave forecast windows for coastal survey or dredging campaign planning · Bridge clearance and fairway depth checks for barge or crane access routes (fairway_nl)
Audience llm-inferred
coastal and marine engineers · hydrologists · civils and BIM engineers planning marine logistics · GIS analysts

Limitations

Constraints llm-inferred
Built and documented for recreational sailing, not engineering or regulatory use · Official charts, almanacs and notices must be checked before relying on outputs · Windy, WorldTides, Stormglass and AISStream need user-supplied API keys; keyed tools only appear when the key is present · AISStream is community-fed and sparse mid-ocean · Some bridges may show '(name unknown)' where OSM has no nearby named bridge · All tool times are UTC · Depends on third-party public endpoints (RWS, NDW, OSM/Overpass, aviationweather.gov, Open-Meteo) with no stated SLA · No public hosted endpoint is documented; the user must run or host an instance (docs use a 'your-host' placeholder) · ENTRY metadata lists OAuth, but DOCS describe API-key URL parameters and state 'No OAuth required' for Claude.ai · Likely single-maintainer personal project ('Built by @ai_kulikov'), so maintenance risk

Some details auto-extracted · unverified

Globalhealthy

mcp
72/100Currency B

Civil3D-mcp ↗

Civil3d MCP server - 180 tools , Knowledge Base, Control UI, Automation

What's in it

Data offered docs-verified
Alignments, profiles, corridors, cross-sections and superelevation · Surface analysis: elevation bands, slope distribution, aspect, watershed, cut/fill volumes · Pipe and pressure network design and hydraulic analysis · Hydrology: flow path tracing, watershed delineation, Rational Method runoff, time of concentration · Grading: feature lines, grading groups, grading criteria, surface generation · Quantity takeoff (earthwork, corridor materials, pipe lengths, parcel areas) and cost estimation · COGO/survey: inverse, traverse, curve solve, survey figures, LandXML import · Plan production: sheet sets, plan/profile sheets, PDF export
Coverage llm-inferred
Global
Formats docs-verified
MCP stdio · HTTP (local, port 3000) · CSV · PDF
Access docs-verified
Open source under MIT licence; requires a locally installed, licensed Autodesk Civil 3D 2026 with API references; optional HTTP bridge credential via MCP_HTTP_TOKEN (entry metadata)

Useful for

Use cases llm-inferred
Cut/fill earthwork volumes between existing ground and finished grade surfaces · Watershed delineation and Rational Method runoff checks for site drainage · Pipe network layout and hydraulic analysis for drainage design · Plan/profile sheet production and PDF publishing for design submissions
Audience llm-inferred
civil/highway design engineers · drainage engineers · hydrologists · BIM/civils engineers

Limitations

Constraints docs-verified
Requires a licensed local Civil 3D 2026 installation and Autodesk API references; Autodesk binaries are not shipped with the repo · Setup docs are Windows-focused (PowerShell commands, Civil 3D plugin built with .NET 8 on Windows) · The Civil 3D plugin must be loaded in a running Civil 3D session for drawing operations · Default MCP surface exposes 34 tools; the full 219-tool surface needs CIVIL3D_ENABLE_TOOL_ALIASES enabled · Entry lists 180 tools, which does not match the 34 public / 219 registered counts in the docs · Destructive or file-writing actions require the preview, approval-token, retry flow · Autodesk-hosted tutorial video playback requires internet access

Globalhealthy

mcp
72/100Currency B

civil3d-mcp ↗

AI-Powered Civil 3D Modeling

What's in it

Data offered docs-verified
AI-generated C# code execution inside Civil 3D (write access, transaction committed) · Read-only C# code queries against the active drawing · Browsable skills library of C# code templates (surfaces, alignments, points, geometry, drawing info, workflows)
Formats docs-verified
MCP stdio · JSON
Access docs-verified
Open source under the MIT licence. Requires a local Autodesk Civil 3D installation, with DLLs copied from Civil 3D and the plugin loaded in the application. No credentials documented.

Useful for

Use cases llm-inferred
Read-only audit of surfaces, alignments and COGO points with their layers before issuing a drawing · Scripted extraction of surface and alignment station/offset data for a drainage or highway design check · Batch creation of repetitive civils objects from skill templates in a design workflow
Audience llm-inferred
civils BIM engineers · drainage engineers · highway engineers · CAD technicians

Limitations

Constraints llm-inferred
Requires Autodesk Civil 3D to be running with the plugin loaded (NETLOAD, C3DMCPSTATUS to verify) · civil3d_execute modifies the drawing and commits the transaction · Roslyn sandbox blocks process execution, file deletion, network requests, registry access and dynamic assembly loading · Default execution timeout is 120000 ms · Argument names for the three tools are not documented in DOCS; the code parameter name in example_tool_calls is inferred · Code is AI-generated and runs with full Civil 3D API access, so output needs review before use on live drawings · Windows and Civil 3D version requirements are not stated in DOCS

Globalhealthy

mcp
70/100Currency C

ArcGIS-Pro-Bridge-MCP-Server ↗

ArcGIS Pro Bridge MCP Server 是一个让 AI 客户端连接本机 ArcGIS Pro 的 MCP Server。

What's in it

Data offered docs-verified
live map and layer listing from the current ArcGIS Pro project · layers with broken data sources · layouts, map frames and default-map candidates from .aprx projects · feature classes, fields and spatial reference from .gdb · generated and executed ArcPy scripts · ArcPy geoprocessing such as Buffer and Clip
Formats docs-verified
MCP stdio
Access llm-inferred
Local install only; requires Windows with ArcGIS Pro installed. No credentials documented in README. MIT licence (catalogue metadata).

Useful for

Use cases llm-inferred
Flag broken layer data sources in a client's ArcGIS Pro project before issuing mapping outputs · Inventory feature classes, fields and spatial reference in a .gdb before data handover or reuse · Run scripted ArcPy buffer or clip of site or watercourse layers for a constraints map · Check layouts and map frames in an .aprx project before reissuing a report figure set
Audience llm-inferred
GIS analysts · civils and drainage engineers working from ArcGIS Pro data · environmental consultants

Limitations

Constraints docs-verified
Windows only · Requires ArcGIS Pro to be installed and launchable; arcpy is not installed via uv · Not for publicly exposed remote GIS services · Not for unattended direct edits to production data · Not a replacement for the ArcGIS Pro GUI and not a general file-system tool · ArcGISProject('CURRENT') may fail to attach; pass the .aprx path instead · Lock errors if layers are being edited or data is in use by ArcGIS Pro · execute_arcpy_code runs code on the local machine; use only on trusted computers · Long or large jobs may time out; narrow scope or raise timeout_seconds · Stdio server is launched by the MCP client, not run manually as a desktop app

Globalhealthy

mcp
70/100Currency A

CAD2GIS ↗

Evidence-first DWG/DXF to GIS conversion, QGIS delivery and CAD MCP tools for AI agents.

Some details auto-extracted · unverified

Globalhealthy

mcp
70/100Currency B

citysense ↗

CitySense is an open-source Python library that bridges geospatial urban data with large language model (LLM) toolchains.

What's in it

Data offered docs-verified
OpenStreetMap data · Mapillary imagery · Sentinel-2 imagery
Coverage docs-verified
Pilot countries: Azerbaijan, Finland, Sweden, Denmark, Norway
Formats docs-verified
Python package (PyPI)
Access llm-inferred
Open source under EUPL-1.2; API credentials required per ENTRY metadata (CDSE client ID/secret, Mapillary access token)

Useful for

Use cases llm-inferred
Urban baseline assembly from OpenStreetMap and Sentinel-2 for a site appraisal · Street-level imagery review of existing highway or drainage assets via Mapillary · LLM-assisted querying of geospatial urban datasets for desk-study reporting
Audience llm-inferred
GIS analysts · civils/BIM engineers · planners

Limitations

Constraints llm-inferred
ENTRY metadata lists no tools, so MCP tool names cannot be documented · DOCS describe a Python toolkit; MCP transport is not stated · Pilot-country focus in DOCS; UK coverage is not stated · Requires Copernicus CDSE and Mapillary credentials per ENTRY metadata · DOCS state no data resolution, update cadence or time span

Some details auto-extracted · unverified

Globalhealthy

mcp
70/100Currency A

ifc-ruhsat ↗

Yapı ruhsatı IFC modelini Türkiye dijital proje yönetmeliğine (RG 33331) göre kontrol eder — CLI + MCP + IDS. Checks IFC models against Türkiye's building-permit BIM regulation.

Some details auto-extracted · unverified

Globalhealthy

mcp
70/100Currency C

metashape-mcp ↗

AI-powered photogrammetry automation — MCP server for Agisoft Metashape Professional 2.3+. Control the full pipeline (alignment, dense cloud, mesh, texture, DEM, orthomosaic, export) using natural language through Claude or any MCP-compatible AI assistant.

What's in it

Data offered docs-verified
DEM generation · orthomosaic creation · dense point cloud · ground classification · mesh generation · texture mapping · contour lines · GCP and coded-marker workflows · CRS/EPSG coordinate setting and reprojection
Coverage llm-inferred
Global
Formats docs-verified
OBJ · PLY · FBX · LAS · LAZ · GeoTIFF · Cesium 3D Tiles · MCP streamable HTTP · MCP stdio
Access docs-verified
MIT licence (free software); requires an Agisoft Metashape Professional 2.3+ installation with Python 3.12 scripting; no credentials documented

Useful for

Use cases llm-inferred
DEM and orthomosaic production from drone survey photos for site topographic baselines · Ground-classified point clouds and surface models from UAV photogrammetry for earthworks or cut-fill inputs · Survey-grade models controlled with ground control points and coded markers · Batch export of DEM, orthomosaic and point cloud deliverables as GeoTIFF or LAZ for GIS and BIM handover
Audience llm-inferred
drone survey and photogrammetry consultants · GIS analysts · civils and BIM engineers · terrain and heritage 3D modellers

Limitations

Constraints llm-inferred
Runs inside a licensed desktop Metashape Professional 2.3+ session; not a standalone data source · Viewport and screenshot tools are unavailable in headless mode · Direct HTTP connection from Claude Code times out after about 60 seconds; the stdio proxy is needed for long jobs and requires a separate FastMCP install · ENTRY tool names (aerial_survey_pipeline, close_range_pipeline, batch_export, diagnose_alignment, optimize_quality_settings) match the docs' guided workflow prompts, not documented callable tools · Tool argument schemas are not documented in the DOCS text · No credentials documented in README; the Metashape licence is separate from the MIT-licensed server

Globalhealthy

mcp
70/100Currency A

navra ↗

Secure MCP gateway for AI agents — deny-wins ACLs, IFC, in-process ML safety, 138 Kani proofs

Some details auto-extracted · unverified

Globalhealthy

mcp
69/100Currency C

CLI-Anything-Arcgis-Pro ↗

Agent-native CLI + live-Pro MCP bridge for ArcGIS Pro — the closed-source counterpart to CLI-Anything's QGIS harness

What's in it

Data offered docs-verified
layout export to PDF · Map Series / map books export · layer attribute query returned as structured rows · run any ArcToolbox geoprocessing tool, with outputs added to the live map · read open project (maps, layouts, active view) · zoom active map to a layer or selection · graduated and unique-value symbology (headless CLI)
Formats docs-verified
JSON · PDF · ArcGIS Pro project (.aprx) · file geodatabase
Access docs-verified
Requires a licensed ArcGIS Pro install (provides ArcPy). Source-available under PolyForm Noncommercial 1.0.0: free for noncommercial use; commercial use needs a paid licence (see LICENSE-COMMERCIAL.md).

Useful for

Use cases llm-inferred
Export print-quality layouts and Map Series from an ArcGIS Pro project for consultation or planning reports · Run clip, buffer or intersect geoprocessing on site or catchment layers and add outputs to the live map for review · Query layer attributes to build constraint or receptor tables for an assessment · Let an agent assemble and symbolise a thematic map in the open project while the consultant watches
Audience llm-inferred
GIS analysts · flood risk consultants · drainage engineers · planners · ecologists

Limitations

Constraints docs-verified
Requires a licensed ArcGIS Pro install; no ArcPy outside Pro · Live bridge needs the .NET add-in built and installed in Pro, plus MCP registration · Verified only on ArcGIS Pro 3.4 / ArcPy 3.4.3 / .NET 8 · Cartographic authoring (symbology, layout, layer management) is still an open frontier on the roadmap · Noncommercial licence; commercial use needs a paid licence · Tool arguments and parameter names are not documented in the README

Globalhealthy

mcp
69/100Currency A

dwg-mcp ↗

MCP gateway for AutoCAD 2022-2027 - text reading, spatial clustering, translation rewriting, and code execution.

What's in it

Data offered llm-inferred
DWG text reading and selection · layer listing and creation · entity query, create, move, copy, and edit · block attribute read and write · linear, aligned, radial, and diameter dimensions · view zoom and capture · DXF export · C# code execution inside the AutoCAD session
Formats docs-verified
MCP stdio
Access llm-inferred
Apache-2.0 licence; local install on Windows with a licensed AutoCAD 2022-2027 session; token auth between server and plugin

Useful for

Use cases llm-inferred
Translate and rewrite foreign-language notes and callouts in place on construction drawing sheets · Batch-update layers, colours and text annotations across a drawing set · Export DWG drawings to DXF for downstream CAD or GIS exchange · Read block attributes and layer structure to audit drawing standards
Audience llm-inferred
BIM/civils engineers · CAD technicians · drainage engineers · GIS analysts

Limitations

Constraints llm-inferred
Windows-only installer (PowerShell, %APPDATA% and %LOCALAPPDATA% paths) · Requires an AutoCAD session running with the plugin loaded; not standalone · Only AutoCAD 2022-2027 is supported; other years need the plugin built from source · The v1.0.0 ZIP ships only the 2024 and 2027 plugin years · dwg_run_lisp refuses every request; dwg_send_code accepts inline synchronous snippets only · Installer requires AutoCAD to be closed · Health and update-check features are marked Unreleased · Usage evidence is limited to 220 tool calls over 19 days on one project

Some details auto-extracted · unverified

Globalhealthy

mcp
69/100Currency C

ifc-bonsai-mcp ↗

An MCP server that connects LLMs to IFC-based BIM workflows in Blender using the Bonsai add-on.

What's in it

Data offered docs-verified
IFC element creation and editing (walls, doors, roofs, stairs) within Blender · IFC scene and object inspection · Trimesh-based procedural geometry converted to IFC elements · IFC/IfcOpenShell knowledge search (RAG)
Formats docs-verified
IFC
Access llm-inferred
Open source, MIT licence (per metadata); no credentials documented in README. Requires a local install of Blender 4.4+ with the Bonsai add-on 0.8.2+ and Python 3.10+.

Useful for

Use cases llm-inferred
Parametric wall, slab and roof creation in an IFC model from natural-language prompts · Scene and object property review of an IFC model via get_ifc_scene_overview and get_object_info · Converting Trimesh-generated custom geometry into native IFC elements for civils BIM models · Screenshot-based QA of BIM model state in Blender
Audience llm-inferred
BIM/civils engineers · BIM coordinators · architects

Limitations

Constraints llm-inferred
Requires a local Blender 4.4+ installation with the Bonsai add-on and the BlenderMCP add-on connected · An empty IFC file must be saved before operations, otherwise changes do not reach the Blender scene · The general execute_code tool is unreliable for IFC operations and may be disabled · Knowledge base features need a local ChromaDB index and an embedding server · The author points users to a newer repository (bonsai-mcp) with a stronger code-generation focus, so this repo may receive less maintenance · Not an official Blender, Bonsai or IfcOpenShell project

Globalhealthy

mcp
69/100Currency C

revit-mcp-server ↗

mcp-servers-for-revit — 80+ AI automation tools for Autodesk Revit via Model Context Protocol (Revit 2023-2026)

What's in it

Data offered docs-verified
project info · model analysis and health checks · clash detection · element create/modify/delete · batch operations · data export
Coverage llm-inferred
Global (ENTRY regions); operates on Revit project models, not a geographic dataset
Formats llm-inferred
MCP stdio · TCP/JSON-RPC 2.0 (plugin link) · JSON · CSV · TXT · PDF · DWG · IFC · RFA (family load input)
Access llm-inferred
Revit plugin requires local Revit install on Windows; MCP tools need no API key as documented; Anthropic API key is optional and only for the built-in chat panel; licence MIT (ENTRY metadata)

Useful for

Use cases llm-inferred
Clash detection between structural and services models before design freeze · Model health audit and warnings review before issuing a model to a client · Material takeoff quantities for early cost estimates · Batch export of sheets to PDF, DWG or IFC for coordination submissions
Audience llm-inferred
BIM/civils engineers · BIM coordinators · structural engineers

Limitations

Constraints llm-inferred
Windows only: Revit runs only on Windows; macOS/Linux are not supported · Operates on the active document only; background documents are not accessible · Revit must be running with the TCP server started via Revit MCP Switch · TCP port 8080 must be free; the server will not start if it is occupied · No undo integration; AI operations are not grouped into a single undo step · send_code_to_revit may fail on assembly conflicts from third-party addins · Parameter names are localised; use BuiltInCategory names for categories · Requires Node.js 18+ for the MCP server · DOCS state both 138 and 124 tools; counts are unreconciled · Fork of an upstream project; maintenance risk

Globalhealthy

mcp
69/100Currency A

vwx-mcp ↗

Vectorworks 2026 MCP server — 248 tools + vs.* knowledge index (3071 signatures). True background writes via native C++ palette (no focus, crash-proof, auto-dismissing error dialogs). Criteria-driven reports, IFC bulk classification, GIS coordinate engine. FastMCP + file-IPC.

What's in it

Data offered docs-verified
248 Vectorworks 2026 MCP tools · 3071-function vs.* knowledge index (signatures)
Formats docs-verified
MCP streamable HTTP
Access docs-verified
Free, open source (MIT); requires a local Vectorworks 2026 session on Windows with the bridge plugin installed

Useful for

Use cases llm-inferred
Batch-editing IFC property sets and bulk classification of Vectorworks BIM models · Generating criteria-driven schedules and reports from a Vectorworks drawing for civils or drainage deliverables · Georeferencing a drawing and converting lat/lon to drawing coordinates via EPSG codes for site layouts
Audience llm-inferred
BIM and civils engineers using Vectorworks 2026 · GIS analysts · CAD/BIM technicians scripting Vectorworks

Limitations

Constraints docs-verified
Every vs.* call runs on the Vectorworks main thread, so long operations block the UI · Export/import verbs (e.g. export_pdf, import_dwg) open Vectorworks modal settings dialogs with no headless path · File-IPC path is Windows-only; macOS/remote use needs the legacy TCP bridge (VWX_TRANSPORT=tcp) · Bridge runs only while the Vectorworks palette is open; closing or pausing it stops the bridge · Multi-step one-time Windows setup: native C++ palette build, menu command and shortcut

Example tool calls docs-verified

Check that the MCP bridge is running and reachable from the client

{
  "tool": "ping",
  "arguments": {}
}

Some details auto-extracted · unverified

Globalhealthy

mcp
68/100Currency D

autocad-mcp ↗

MCP server for AutoCAD LT v3.1: freehand AutoLISP execution, 8 consolidated tools, File IPC + ezdxf backends, focus-free dispatch, undo/redo, P&ID symbols, and robust IPC with ESC prefix and UTF-8 fallback. Companion agent skill: puran-water/autocad-drafting

What's in it

Data offered docs-verified
DXF export · PDF plotting · PNG view screenshots · entity create/modify (lines, circles, polylines, hatches, MText) · layer management · block insert and attribute read/update · P&ID symbol insertion (CTO library) · undo/redo · arbitrary AutoLISP execution (File IPC only)
Formats docs-verified
DXF · PDF · PNG · MCP stdio
Access docs-verified
Open source, MIT licence, no credentials documented. File IPC backend requires AutoCAD LT 2024+ on Windows; ezdxf headless backend runs without AutoCAD on any platform.

Useful for

Use cases llm-inferred
Generating headless DXF drawing outputs for drainage or civils sheets on machines without AutoCAD (ezdxf backend) · Batch-updating title-block or equipment tag attributes across a set of drawings · Drafting P&ID-style process schematics for water treatment schemes with CTO symbols · Exporting drawings to DXF and PDF for issue to design teams or clients
Audience llm-inferred
CAD technicians · drainage engineers · civils and BIM engineers · process engineers drafting P&IDs

Limitations

Constraints docs-verified
File IPC backend needs Windows 10/11, native Windows Python 3.10+, and AutoCAD LT 2024 or newer running · AutoLISP is not supported on AutoCAD LT for Mac · offset, fillet and chamfer are File IPC only · execute_lisp is File IPC only · plot_pdf and undo/redo are not available on the ezdxf backend · P&ID symbol insertion requires the CTO P&ID Symbol Library installed at C:\PIDv4-CTO\ · Beta-style v3.1 release with IPC timing dependencies (timeout 1-300 s)

Example tool calls docs-verified

Confirm which backend is active (file_ipc if AutoCAD is running, ezdxf otherwise)

{
  "tool": "system",
  "arguments": {
    "operation": "status"
  }
}

Export the current drawing as DXF (path argument not documented, so omitted)

{
  "tool": "drawing",
  "arguments": {
    "operation": "save_as_dxf"
  }
}

Capture the current AutoCAD view as a PNG for visual checking

{
  "tool": "view",
  "arguments": {
    "operation": "get_screenshot"
  }
}

Globalhealthy

mcp
68/100Currency A

fancy-map ↗

Engine-agnostic Map component for the Fancy UI suite — OpenStreetMap (Leaflet) + Google Maps providers, live tracking, Human+ MCP bridge.

Some details auto-extracted · unverified

Globalhealthy

mcp
68/100Currency A

horizun-msproject-mcp ↗

MCP server for Microsoft Project: no Java, no Project install. Critical-path engine, DCMA 14-point assessment, verified writes, and a BIM 4D bridge.

What's in it

Data offered docs-verified
critical path · float distribution · driving path · day-by-day resource overallocation · DCMA-14 assessment · earned value (BCWS/BCWP/ACWP, SPI, CPI, EAC, TCPI) · quantitative schedule risk (Monte Carlo P10/P50/P80/P90) · period status report · dry-run schedule change impact
Coverage llm-inferred
Global
Formats docs-verified
MPP · Primavera XER · MSPDI XML · PMXML · CSV · JSON · Power BI dataset · MCP stdio
Access docs-verified
Open source, MIT licence (per entry metadata); no Microsoft Project, JVM or paid licence required; installed as a .NET tool

Useful for

Use cases llm-inferred
DCMA-14 schedule quality check on a civils programme before issue · P80 schedule risk contingency from Monte Carlo analysis of a construction programme · Period earned value and status reporting (SPI, CPI, EAC, forecast finish) against a baseline · Linking schedule tasks to BIM model elements by shared code for 4D sequencing
Audience llm-inferred
project controls and PMO staff · programme managers · planners and scheduling engineers · BIM/civils engineers

Limitations

Constraints docs-verified
write_native_mpp is unavailable on the file backend; without Microsoft Project the save falls back to MSPDI · level_resources refuses on the file backend rather than approximating Microsoft Project's leveller · Moving a task to another position in the outline is not offered · Tool schemas add about 13,500 tokens to every prompt while the server is connected

Some details auto-extracted · unverified

Globalhealthy

mcp
68/100Currency A

new-acad ↗

AI bridge for AutoCAD / Civil 3D: in-app AI palette + 270 MCP tools + 390+ plugin methods - surfaces, alignments, corridors, pipe networks, grading, batch entity editing, file channel. | 用自然语言操作 AutoCAD / Civil 3D

What's in it

Data offered docs-verified
CAD entity read/edit (layers, blocks, properties, batch handles) · Civil 3D surfaces (create, bulk point add) · Civil 3D alignments, profiles and sheets · Corridors and corridor surfaces · Grading optimisation (design surface to TIN) · Pipe networks, parcels, catchments, intersections · Corridor earthwork volumes · Site mass haul (cut/fill balance, haul distances) · Quantity and cost estimates · Bitmap-to-vector tracing · Office/zip file reading (docx, xlsx, pptx, zip, doc, xls)
Formats docs-verified
PNG · BMP · JPEG · GIF · TIFF · docx · xlsx · pptx · zip · doc · xls · MCP (HTTP POST :3000/execute) · TCP JSON-RPC 2.0
Access docs-verified
Open source, MIT licence (code). Requires a user-supplied, licensed AutoCAD or Civil 3D 2025/2026 on Windows. Default LLM mode needs the user's own LLM API key.

Useful for

Use cases llm-inferred
Cut/fill earthwork estimate for a linear corridor scheme (computeCorridorEarthwork) · Draft site grading and cut/fill balance for a development platform (optimizeGrading) · Batch editing of layers, blocks and entities across a drawing from natural language · Converting a raster site plan or sketch into CAD linework for further design
Audience llm-inferred
civil design engineers using Civil 3D · drainage engineers · BIM and civils CAD technicians

Limitations

Constraints docs-verified
Windows only · Needs a licensed AutoCAD/Civil 3D session open; plugin DLL is locked while C3D runs · Not an Autodesk product and not endorsed by Autodesk · Docs and default UI are Chinese; English UI added in v1.4.0 · No UK datasets, standards or geographic coverage stated · Docs cite both 270 and 274 MCP tools, so tool count may be inconsistent · 12 methods use a command channel with a 25 s timeout; some C3D features are not exposed to the AI · Corridor regeneration can take 30 s to 10 min and queues other requests (HOST_BUSY) · Grading optimisation is a draft generator; multi-level walls and pipe avoidance need manual review · Not a data API: no hydrology, gauge or DEM feeds documented

Example tool calls docs-verified

Get a budgeted summary of the open drawing (layers, spatial grid, entity counts). Argument schema not documented; docs say fullResult:true returns the full result.

{
  "tool": "summarizeDrawing",
  "arguments": {}
}

Compute corridor cut/fill earthwork volumes. Docs report about 0.1% difference from the C3D surface volume method. Argument schema not documented.

{
  "tool": "computeCorridorEarthwork",
  "arguments": {}
}

Generate a draft design surface from slope limits, platform levels, walls and steps, with cut/fill balance. Argument schema not documented.

{
  "tool": "optimizeGrading",
  "arguments": {}
}

Some details auto-extracted · unverified

Globalhealthy