AutoCAD MCP Pro

Production-grade AutoCAD automation for AI agents. Live through COM on Windows, or headless through ezdxf anywhere — one typed contract, two engines.

Install · Tools · Engines · Evidence · Limits · Config · Changelog

_{Not a mockup. Every line on this sheet was drawn by the tools this server exposes — ISO layers, involute gear geometry, DIN 6885 keyway, section A-A, ISO 129 dimensions, an ISO 286 H7 bore fit, ISO 7200 title block — then rendered headlessly by view_screenshot. drawing_critique returns 0 issues on it. Rebuild it with python scripts/render_readme_showcase.py.}

_{The same server as a P&ID drafter. pid_from_spec placed five catalogue symbols as real blocks with TAG attributes and named ports, routed four port-to-port lines with ISA-5.1 classes and line numbers in one transaction, then pid_graph read the sheet back: 5 nodes, 4 edges, 0 dangling ends, confidence_min 1.0. Rebuild it with python scripts/render_readme_pid.py.}

_{The catalogue, authored in code. 43 ISO 10628-2 / ISA-5.1 symbols — 11 valve bodies × 6 actuators, 6 rotating machines, 4 heat exchangers, 6 parametric vessels, 7 inline items, the instrument bubble in 4 types × 5 locations, 5 line markers, 3 connectors — every one placed through pid_symbol_insert's own path. Rebuild it with python scripts/render_pid_catalog.py.}

v1.5 release snapshot: 169 tools · 8 resources · 5 prompt templates · 2423 collected tests. 166 is the registered count; a default install advertises 161 over tools/list, because ENABLE_3D is unset. system_about is the runtime authority.

Why this exists

A big MCP server is expensive to be connected to. The full catalog costs a client 44,930 tokens before it has asked for anything. Discovery mode replaces it with two tools and costs 356.

A drafter searches for FILLET, not entity_fillet. Those command names appeared in no tool name or description — df = 0 against a stock index, not badly ranked but absent. The fix was data: an authored corpus of 162 AutoCAD command names and 731 synonym phrases covering all 166 tools. A test refuses to let a tool exist without one.

Advertised surface	Tools seen	Idle cost
TOOL_PROFILE=full (default)	161	44,930 tokens
TOOL_PACKS=core (full profile)	152	41,420 tokens
TOOL_PROFILE=lean	50	14,007 tokens
DISCOVERY_MODE=search	2	356 tokens

Note

Offline ratio estimates, not tokenizer counts, and the uncached ratio — prompt caching amortises the idle term. benchmarks/token_suite.py --tokenizer anthropic counts for real.

Install

pip install autocad-mcp-pro     # or: uvx autocad-mcp-pro
autocad-mcp                     # stdio MCP server, backend auto-selected

AUTOCAD_MCP_BACKEND=ezdxf autocad-mcp    # portable DXF engine, no AutoCAD needed
AUTOCAD_MCP_BACKEND=com   autocad-mcp    # live AutoCAD (needs the [com] extra)

Extra	Pulls in	For
(none)	fastmcp, ezdxf, pydantic	Headless DXF on any OS — no rendering
[com]	pywin32, Pillow	Live AutoCAD control + window capture
[pdf]	matplotlib	PDF export and headless PNG, any platform
[full]	everything above	Development and CI

Note

The bare install cannot draw pixels — ezdxf.addons.drawing imports Pillow unconditionally, so every render path needs it. Add [pdf] for images on Linux or macOS.

Wire it to an MCP client

{
  "mcpServers": {
    "autocad": {
      "command": "autocad-mcp",
      "env": {
        "AUTOCAD_MCP_BACKEND": "auto",
        "ALLOWED_PATHS": "C:\\Users\\you\\Documents\\AutoCAD",
        "TOOL_PROFILE": "full",
        "DISCOVERY_MODE": "off"
      }
    }
  }
}

Claude Desktop, Cursor, or any stdio MCP host. For HTTP: autocad-mcp --transport http --port 8000 — loopback only unless remote HTTP is explicitly enabled and a bearer token is set.

What you get

Area	What it does
Drawing lifecycle	create, open, save, export DXF/PDF, audit (repairs), purge, undo/redo
Geometry	lines, arcs, polylines, splines, hatches, trim/extend/fillet/chamfer, handle-preserving edits
Annotation	ISO 129 toleranced dimensions, ISO 286 fits (fit="H7"), TABLE, MLEADER, GD&T frames and datums (ISO 1101)
Engineering generators	involute gears (front + section A-A), DIN 6885 keyed bores, ISO A3 title block
P&ID	catalogue blocks with ports and tags (ISO 10628-2 / ISA-5.1), port-to-port lines with ISA-5.1 classes and line numbers, pid_graph reads any P&ID back with confidence, instrument index / line list / equipment list, pid_from_spec one-call sheets
Discovery	search_tools ranked over an AutoCAD command and synonym corpus — FILLET, BPOLY, QSELECT, WBLOCK, OVERKILL, CHSPACE each rank #1 of the 161-tool advertised catalog
Batching	cad_batch runs a step list in one round trip; fields= projects 11 result-heavy tools
Paper space	tab lifecycle, viewports, entity_change_space (CHSPACE), drawing_export_pdf(layout=…)
Selection	window vs crossing stated back to the caller; a polygon tested against its own shape, not its bounding box
Boundaries	boundary_trace (BOUNDARY/BPOLY) chains loose edges into one closed polyline, arcs kept as bulges (headless)
Measurement	analysis_measure_entity measures what is in the drawing, by handle
Hatch depth	gradients, in-place edits, typed edge boundaries (headless), island styles
Annotation objects	WIPEOUT, REVCLOUD, MTEXT background masks (headless), text find/replace
3D solids	solid_box/cylinder/extrude/revolve/boolean on live AutoCAD (ENABLE_3D=true)
Quality loop	drawing_preflight → drawing_plan → drawing_critique → drawing_refine → drawing_finalize (0–100 score)
Delivery	drawing_deliver: DXF/PDF/PNG + SHA-256 manifest + reopen-parity checks

_{166 tools in 20 groups; TOOL_PACKS=core hides the nine pid_* tools from a client that never draws a P&ID. Plus 8 resources that cost nothing in the tool budget (autocad://drawing/info, layers, blocks, entities/stats, entities/{layer_name}, system/status, pid/symbols, standards/isa51) and 5 prompt templates.}

Two rules worth knowing. Every coordinate in and out of a tool is WCS on both engines — the one exception is TEXT rotation, which stays in the entity frame because a mirrored TEXT is mirror-imaged and no scalar angle expresses that. And never read vertices back and shoelace them: that loses 28.2% of the area on a semicircular edge, silently. analysis_measure_entity(handle) reads the real geometry and states its own accuracy.

The two engines

One contract across 19 modules in backends/contracts/. @capability(key, reason=…) supplies a default that refuses, and a test holds both backends' capability key sets equal. There are 27 capability keys; read system_capabilities at runtime rather than trusting the table — five of them depend on how the machine is set up.

Capability	COM (live AutoCAD)	ezdxf (headless)
Live document control	✅	—
Cross-platform, no AutoCAD	—	✅
Transactions and rollback	✅	✅
Paper-space layouts + viewports	✅	✅
Viewport model-content rendering	✅	✅ ᵐ (no borders)
Selection window / crossing / polygon	✅	✅
Entity area by handle	ActiveX .Area	Analytic, bulges included
HATCH filled area (islands subtracted)	AutoCAD's own number	Loops walked, hatch_style reported
REGION / 3DSOLID area	✅	— ACIS is opaque to ezdxf
3D solids	✅ with ENABLE_3D=true	—
DWG write	✅	—
WIPEOUT · MTEXT background colour	— verified absent, AutoCAD 2026	✅
REVCLOUD · BPOLY · typed hatch edges	— no ActiveX member	✅
CHSPACE	— unverified on a live seat	✅ ᶜ
Undo history	✅	opt-in — EZDXF_UNDO_DEPTH
TABLE and MLEADER	Native	Portable composite
Screenshots and PDF	Window capture	Matplotlib ᵐ

_{ᵐ Needs matplotlib ([pdf]/[full]); without it png, pdf, viewport_render and handle_overlay report unsupported headlessly.
ᶜ Headless CHSPACE carries four named restrictions in its capability reason: top-view untwisted viewports only, dimensions refused unless frozen, ACIS/proxy/table refused, viewport clipping reported rather than applied.}

Important

Every COM path added in v1.5 was executed against a live AutoCAD 2026. That run found four defects reading could not — the largest being that AcadPViewport has no ViewCenter member, so the line setting it had been silently swallowed by a bare except since v1.4.

Evidence

Self-measurement, produced by scripts in benchmarks/.

Correctness — every release re-proves itself

29 deterministic headless checks against the previous tag and the current tree, each in its own subprocess so a hard crash counts as a miss rather than killing the run.

Version	Checks passing	Pass rate	Fixed	Regressed
v1.5.1 (baseline)	26 / 29	89.7 %	—	—
v1.6.0-dev (this branch)	29 / 29	100 %	3	0

The three fixed rows are the P&ID checks, new capability (miss → pass): block_define round-trips an ATTDEF, pid_tag_parse reads FIC-101, and pid_graph counts the example sheet's four edges. The 26 checks v1.5.1 was gated on all still pass. Against the older v1.4.0 baseline the same 26 reported 21 / 26 → 26 / 26, five fixed, zero regressed — two of those were repaired defects (fail → pass), the diameter and radius callouts, which measured the leader as geometry and dimensioned a 40 mm bore as 60 at default settings.

The task matrix — six tasks that can fail

An earlier matrix scored this server 10/10, which carried no information: every task in it exercised something the server was built around. Five were added in 1.5 because they can fail, and three did while being written; 1.6 adds a sixth that the P&ID reader can fail on its own.

Task	Verified against
tool_discovery	six AutoCAD command names, each ranking #1
token_budget	44,930 → 356 tokens, against a ceiling fixed in advance
hatch_islands	300 filled with the island, 400 ignoring it
selection_filter	window 1, crossing 2, bounding box 3, polygon 1
measure_from_handle	139.2699 against the 100.0 a vertex shoelace gives
pid_roundtrip	the example sheet drawn by pid_from_spec, read back by pid_graph, which never sees the spec: 5 nodes, 4 edges, 0 dangling, confidence_min 1.0, FIC-101 wired to FCV-101

Headless performance

Workloads call the same backend methods the MCP tools call, so server-side overhead is included. Numbers move with hardware; the report records the machine fingerprint. Read the next section before quoting them — this release is slower than v1.4.0 at creating entities, on purpose.

What this release is bad at

A page that only lists strengths is a page that has not been measured.

Entity creation is slower than v1.4.0. Median of three runs on one machine and interpreter: 2,000 lines 1.4× slower, the 10,000-line roundtrip 2.3× slower. Attributed — setting EZDXF_CALL_TIMEOUT=0 returns creation to v1.4.0's numbers. The cost is the per-call asyncio.wait_for wrapped around every headless call so one hung call can no longer wedge a server whose document lock is a single asyncio.Lock. Deliberate trade, documented knob, on the 1.6 roadmap. The premium quality pass went the other way — 3.7× faster.

Wave A shipped 13 tools of a planned 39. Every cut is backed by a measurement. The REGION and 2D-boolean family went because add_region() produces a REGION with zero ACIS bytes, and the greiner_hormann substitute loses 28.2% of the area on a square with one semicircular edge.

system_run_command / system_run_lisp are a guardrail, not a security boundary — the rejection message says so in those words. A 36-verb denylist refuses the obvious cases, but AutoCAD accepts hundreds of commands and DANGEROUS_COMMANDS_ENABLED=true switches it off entirely.

Path validation is per-tool, and unscoped until you scope it. With ALLOWED_PATHS empty — the default — the only positive bound is a ten-entry system-directory denylist that does not include C:/Users, /home, /root or /var. Set ALLOWED_PATHS.

Non-AutoCAD ProgIDs are unverified. CAD_PROGID changes which COM application the backend attaches to; nothing beyond the connection has been tested against BricsCAD, ZWCAD or GstarCAD.

Known limitations of the P&ID track

The router is orthogonal and blind. pid_line_draw picks auto from seven orthogonal candidates — straight, one bend, two bends through the midline, or a stub out of each port joined by three bends — by fewest bends then length (or takes direct, or the waypoints you pass), and counts the lines it crosses — crossings comes back in every result and pid_from_spec sums it — but it does not route around them. Obstacle-avoiding routing is out of 1.6; move the symbol or pass waypoints.

A foreign drawing is read at the confidence it earns, and no higher. A block this server did not place classifies by its TAG-style attribute or a keyword in its name at 0.6 with inferred ports, or at 0.3 when a line merely touches it — and a 0.3 guess can neither raise nor suppress a critique finding. On the live engine an INSERT turned off a right angle whose members ActiveX cannot measure reports an approximate box; pid_graph takes the tighter wall of that and the attribute-inclusive box on every side, but a line ending on such a body's true edge can still be reported dangling with the nearest port as the hint. Read stats.confidence_min first.

Same-file only, DXF headlessly. Off-page connectors link within one drawing; links across files are not resolved. The headless engine reads DXF; a DWG P&ID is read through the live backend. No DEXPI/Proteus export and no ISA-5.2 binary-logic symbols in this release, recorded in the spec so nobody re-derives the cut.

Configuration

Nothing loads a .env file — export these, or set them in your MCP client's env block.

All 17 environment variables

Variable	Default	Purpose
AUTOCAD_MCP_BACKEND	auto	auto, com, or ezdxf
CAD_PROGID	AutoCAD.Application	COM ProgID the live backend attaches to
TOOL_PROFILE	full	lean (50 curated tools) or full
TOOL_PACKS	all	Vertical packs to advertise: core,pid (core always on)
DISCOVERY_MODE	off	search replaces the catalog with search_tools + call_tool
ENABLE_3D	false	Expose the opt-in solid_* tools (COM)
LOG_LEVEL	INFO	Python logging level
ALLOWED_PATHS	(empty)	Comma-separated absolute paths the server may access
MAX_UNDO_STACK	5	Maximum retained undo snapshots
EZDXF_UNDO_DEPTH	0	Headless undo history depth; 0 disables it
MAX_DXF_BYTES	52428800	Reject larger DXF input; 0 disables
MAX_LIST_LIMIT	5000	Bound list/selection response sizes
COM_CALL_TIMEOUT	60	Per-call live AutoCAD timeout (s); 0 disables
EZDXF_CALL_TIMEOUT	120	Per-call headless timeout (s); 0 disables
DANGEROUS_COMMANDS_ENABLED	false	Allow blocked commands/LISP; reported as unsafe mode
ALLOW_REMOTE_HTTP	false	Permit a non-loopback HTTP bind
MCP_AUTH_TOKEN	(empty)	Bearer token required for remote HTTP

Architecture

flowchart LR
  A[MCP host / AI agent] --> B[FastMCP 3 server]
  B --> C[Error · audit · timing · logging · capability-refusal middleware]
  C --> D[Typed contract · 19 modules · 27 capabilities]
  D --> E[COM backend<br/>single-STA thread]
  D --> F[ezdxf backend<br/>asyncio.to_thread]
  E --> G[Live AutoCAD]
  F --> H[Headless DXF]
  E --> I[Engineering · critique · scoring · delivery]
  F --> I

The drawing at the top of this page, as tool calls

The sequence in scripts/render_readme_showcase.py, which produced the hero sheet:

drawing_new()                                    ISO linetypes + layers bootstrapped
drawing_apply_iso_layers("mech")                 ISO 128 lineweights per layer
drawing_settings({units: "mm", linear_precision: 2})

gear_draw_spur_front_view(                       involute flanks, not a decorated circle
    module=6, teeth=24, center=[135, 172],
    bore_diameter=40, keyway_width=12, keyway_depth=3.3)
gear_draw_section_aa(x_offset=300, face_width=46)

dimension_linear(...)                            ISO 129
dimension_diameter(..., fit="H7")                deviations from authored ISO 286 tables
titleblock_apply_iso_a3(title="SPUR GEAR m6 z24", material="C45E", ...)

drawing_critique(focus=None)                     -> []  must be empty before finalize
view_screenshot()                                -> PNG

The full production loop adds drawing_preflight and drawing_plan at the front, drawing_refine after the critique, and layout_create + viewport_create + drawing_deliver at the end.

Development

uv sync --locked --all-extras
uv run pytest
uv run ruff check . && uv run ruff format --check .

uv.lock pins the whole transitive graph, so this reproduces CI exactly. CI runs Linux (3.11 / 3.12), Windows (mocked-COM), plus package, Docker and MCP-registry jobs; the release gate runs the same three test lanes before anything reaches PyPI. Releases are tag-driven: git tag vX.Y.Z && git push origin vX.Y.Z.

Roadmap (1.6)

Give back the creation throughput the per-call timeout costs — arm it only for calls that can actually block. ISO 286 interference shafts r/s/t/u. HATCH boundary geometry in entity_get (1.5 measures a hatch's filled area but does not hand back its loops). REGION/3DSOLID area headlessly, which needs a modelling kernel ezdxf does not have. An allowlist of permitted AutoLISP heads, replacing the denylist — enumerating dangerous symbols does not terminate. ezdxf.recover as a fallback on drawing_open.

Features ship when their contracts and limitations are testable — not when they make a longer checklist.

Star History

Author

Umutcan Edizsalan · Mechanical engineering work at Anka-Makine · GitHub @U-C4N

Built from production drawing work, then made model-agnostic through MCP.

License

MIT

mcp-name: io.github.u-c4n/autocad-mcp