Provides tools for controlling AutoCAD LT, including drawing/file management, entity creation and modification, layer management, block insertion/attribute handling, annotations, P&ID operations, view controls, and system status/execution, with support for both live AutoCAD via File IPC and headless DXF generation via ezdxf.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@AutoCAD MCP Server Create a new drawing, draw a circle, and save as DXF"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP server for AutoCAD LT automation and headless DXF generation.
Two backends, one API:
Backend | Runtime | Requires AutoCAD? | Screenshot |
File IPC | Windows Python | Yes — AutoCAD LT 2024+ (Windows) | Win32 PrintWindow |
ezdxf | Any platform | No (headless) | matplotlib render |
The server exposes 8 consolidated tools (drawing, entity, layer, block, annotation, pid, view, system) over the MCP stdio transport. An MCP client (Claude Desktop, Claude Code, etc.) connects and drives AutoCAD through natural-language requests.
Windows 10/11 (the File IPC backend uses Win32 APIs for focus-free window messaging)
AutoCAD LT 2024 or newer — AutoLISP support was added in LT 2024 for Windows. AutoCAD LT for Mac exists but does not support AutoLISP.
Python 3.10+ (Windows native — not WSL Python)
uv package manager (install guide)
The ezdxf headless backend works on any platform (Linux, macOS, WSL) for offline DXF generation without AutoCAD installed.
Related MCP server: Greenloom CAD MCP Server
git clone https://github.com/puran-water/autocad-mcp.git
cd autocad-mcp
uv syncOpen AutoCAD LT and load mcp_dispatch.lsp using APPLOAD:
Type APPLOAD in the AutoCAD command line
Browse to <repo>/lisp-code/mcp_dispatch.lsp
Click Load
You should see: === MCP Dispatch v3.1 loaded === and Ready for commands via (c:mcp-dispatch)
Tip: Add the file to your AutoCAD Startup Suite (in the APPLOAD dialog) so it loads automatically with every drawing.
Add to your MCP client configuration (e.g. Claude Desktop claude_desktop_config.json):
{
"mcpServers": {
"autocad-mcp": {
"command": "C:\\path\\to\\autocad-mcp\\.venv\\Scripts\\python.exe",
"args": ["-m", "autocad_mcp"],
"env": { "AUTOCAD_MCP_BACKEND": "auto" }
}
}
}Key points:
The command must point to the Windows Python inside the project venv (not WSL python).
AUTOCAD_MCP_BACKEND can be auto (default — tries File IPC, falls back to ezdxf), file_ipc (requires AutoCAD), or ezdxf (headless only).
If your MCP client runs in WSL (e.g. Claude Code), launch the server through cmd.exe so it runs as a native Windows process:
{
"mcpServers": {
"autocad-mcp": {
"type": "stdio",
"command": "cmd.exe",
"args": ["/d", "/s", "/c", "cd /d C:\\path\\to\\autocad-mcp && .venv\\Scripts\\python.exe -m autocad_mcp"],
"env": { "AUTOCAD_MCP_BACKEND": "auto" }
}
}
}From your MCP client, call:
system(operation="status")You should see backend: "file_ipc" if AutoCAD is running, or backend: "ezdxf" for headless mode.
drawing — File/drawing managementOperation | Description | File IPC | ezdxf |
| Reset to clean drawing (erase all + purge) | Yes | Yes |
| Open an existing drawing | Yes | Yes (DXF) |
| Get entity count and layers | Yes | Yes |
| Save current drawing (to path if given) | Yes | Yes |
| Export as DXF | Yes | Yes |
| Plot to PDF | Yes | No |
| Purge unused objects | Yes | Yes |
| Get system variables by name | Yes | Yes |
| Undo last operation | Yes | No |
| Redo last undone operation | Yes | No |
entity — Entity CRUD + modificationCreate: create_line, create_circle, create_polyline, create_rectangle, create_arc, create_ellipse, create_mtext, create_hatch
Read: list, count, get
Modify: copy, move, rotate, scale, mirror, offset*, array, fillet*, chamfer*, erase
*
offset,fillet,chamferare File IPC only (not supported in ezdxf headless backend).
layer — Layer managementlist, create, set_current, set_properties, freeze, thaw, lock, unlock
block — Block operationsOperation | File IPC | ezdxf |
| Yes | Yes |
| Yes | Yes |
| Yes | Yes |
| Yes | Yes |
| Yes | Yes |
| No | Yes |
annotation — Text, dimensions, leaderscreate_text, create_dimension_linear, create_dimension_aligned, create_dimension_angular, create_dimension_radius, create_leader
pid — P&ID operations (CTO symbol library)setup_layers, insert_symbol, list_symbols, draw_process_line, connect_equipment, add_flow_arrow, add_equipment_tag, add_line_number, insert_valve, insert_instrument, insert_pump, insert_tank
P&ID symbol insertion requires the CAD Tools Online (CTO) P&ID Symbol Library installed at
C:\PIDv4-CTO\. The ezdxf backend has built-in CTO library support. For the File IPC backend, some P&ID operations require additional LISP helpers — see the P&ID section in the wiki for setup details.
view — Viewport and screenshotOperation | Description |
| Zoom to show all entities |
| Zoom to a specified window |
| Capture current AutoCAD view as PNG |
Screenshots use PrintWindow (Win32) for the File IPC backend — works even when AutoCAD is minimized or in the background. The ezdxf backend renders via matplotlib.
system — Server managementstatus, health, get_backend, runtime, init, execute_lisp
execute_lispruns arbitrary AutoLISP code (File IPC only). Passdata: {code: "(+ 1 2)"}. This turns the server into an extensible automation platform — any valid AutoLISP expression can be executed.
MCP Client (Claude)
│ stdio (JSON-RPC)
▼
Python MCP Server (autocad_mcp)
│
├── File IPC Backend ──► C:/temp/*.json ──► mcp_dispatch.lsp (AutoCAD LT)
│ PostMessageW(WM_CHAR) to MDIClient — no focus steal
│
└── ezdxf Backend ──► in-memory DXF (headless, no AutoCAD needed)The File IPC backend sends keystrokes to AutoCAD's MDIClient window via PostMessageW(WM_CHAR), triggering the (c:mcp-dispatch) AutoLISP command. This approach does not steal window focus — you can continue working in other applications while automation runs.
Variable | Default | Description |
|
| Backend selection: |
|
| Directory for IPC command/result JSON files (must match on both Python and LISP sides) |
|
| IPC command timeout in seconds (1-300) |
|
| Disable screenshot capture (text feedback only) |
Note: If you change
AUTOCAD_MCP_IPC_DIR, you must also update the*mcp-ipc-dir*variable inmcp_dispatch.lspto match.
uv sync
uv run pytest tests/ -vAutoLISP was added to AutoCAD LT in the 2024 release (Windows only). AutoCAD LT for Mac does not support AutoLISP.
Supported (LT 2024+ Windows) | Not Supported |
| VLIDE (Visual LISP IDE) |
All |
|
File I/O ( | Express Tools |
Entity access ( | 3D operations |
Selection sets | AutoLISP on Mac |
The mcp_dispatch.lsp dispatcher is fully compatible with LT 2024+.
execute_lisp — Run arbitrary AutoLISP code via temp file pattern. Turns the server from a fixed command set into an extensible automation platform.
Undo / Redo — Single-step undo and redo via drawing tool.
Drawing open — Open existing .dwg files programmatically (FILEDIA suppressed).
Drawing create — Now resets current drawing (erase all + purge) instead of _.NEW, preserving the LISP dispatcher namespace.
Drawing save with path — save with a path parameter uses SAVEAS; without path uses QSAVE.
get_variables fix — Respects the names parameter; returns requested variables with proper type handling.
Polyline/leader fix — Point arrays properly encoded via semicolon-delimited format.
ESC prefix — Sends 2x ESC before each dispatch to cancel stale pending commands from prior timeouts.
UTF-8/cp1252 fallback — Handles non-ASCII characters in LISP result files (AutoCAD writes Windows-1252).
Configurable IPC timeout — AUTOCAD_MCP_IPC_TIMEOUT env var (1–300 seconds, default 10).
Thread-safe backend init — asyncio.Lock prevents parallel initialization races.
MIT
Annotation: text, dimensions, and leaders.
Operations: create_text — data: {x, y, text, height?, rotation?, layer?} create_dimension_linear — data: {x1, y1, x2, y2, dim_x, dim_y} create_dimension_aligned — data: {x1, y1, x2, y2, offset} create_dimension_angular — data: {cx, cy, x1, y1, x2, y2} create_dimension_radius — data: {cx, cy, radius, angle} create_leader — data: {points: [[x,y],...], text}
| Name | Required | Description | Default |
|---|---|---|---|
| data | No | ||
| operation | Yes | ||
| include_screenshot | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description goes beyond the readOnlyHint:false annotation by exposing the specific operations and their data shapes, implying mutating behavior. However, it does not disclose side effects, required permissions, or what happens on success/failure, leaving behavioral transparency only partially covered.
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, front-loads the purpose, and organizes operations as a scannable list. It is not overly verbose, though the list structure could be improved with a short introductory verb phrase.
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 has six distinct operations and an output schema, the description covers the operation specifics but lacks broader context: no use-case examples, no prerequisites, and no explanation of when each dimension type is appropriate. The output schema, if present, may cover return values, but the description remains incomplete for a tool with this complexity.
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?
The schema has 0% description coverage, but the description compensates well by detailing the data fields for each operation and implicitly enumerating valid operation values. Missing is any explanation of the include_screenshot parameter, which is a clear gap.
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 deals with annotations (text, dimensions, leaders) and enumerates specific create operations, so an agent can infer the tool creates annotation objects. However, it does not explicitly distinguish this from sibling tools like entity or drawing, and the opening line is more of a category label than a verb-driven statement.
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?
There is no guidance on when to use this tool versus alternatives, no mention of prerequisites (e.g., open drawing), and no exclusionary context. The description simply lists operations without explaining the appropriate scenarios for each.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Block definition, insertion, and attribute management.
Operations: list — List all block definitions. insert — data: {name, x, y, scale?, rotation?, block_id?} insert_with_attributes — data: {name, x, y, scale?, rotation?, attributes: {tag: value}} get_attributes — data: {entity_id} update_attribute — data: {entity_id, tag, value} define — data: {name, entities: [{type, ...}]}
| Name | Required | Description | Default |
|---|---|---|---|
| data | No | ||
| operation | Yes | ||
| include_screenshot | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description aligns with annotations (readOnlyHint: false) by showing mutating operations like insert and update_attribute. It does not, however, disclose potential side effects, error conditions, or prerequisites beyond the operations themselves.
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 reasonably concise, front-loading the purpose and using a structured bullet list for operations. It avoids unnecessary prose, though the operation list is somewhat lengthy.
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?
The description covers all primary operations and their data requirements, which is sufficient given the presence of an output schema and annotations. It could be enhanced by explaining the 'include_screenshot' flag and any return values, but overall it provides a complete picture for this multi-operation tool.
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?
The input schema has three parameters with zero description coverage. The description compensates by listing valid operation values and providing the expected shape of the 'data' object for each operation, though it omits the 'include_screenshot' parameter.
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's scope: block definition, insertion, and attribute management. It enumerates specific operations (list, insert, define, etc.), making the purpose distinct from sibling tools like entity or layer.
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 list of operations effectively tells an agent when to use this tool (e.g., for block insertion or attribute updates). However, it does not explicitly mention when not to use it or name alternative tools for non-block tasks.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Drawing file management.
Operations: create — Create a new empty drawing. data: {name?} open — Open an existing drawing. data: {path} info — Get drawing extents, entity count, layers, blocks. save — Save current drawing. data: {path?} (saves to path if given, else QSAVE) save_as_dxf — Export as DXF. data: {path} plot_pdf — Plot to PDF. data: {path} purge — Purge unused objects. get_variables — Get system variables. data: {names: [...]} undo — Undo last operation. redo — Redo last undone operation.
| Name | Required | Description | Default |
|---|---|---|---|
| data | No | ||
| operation | Yes | ||
| include_screenshot | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The annotation readOnlyHint=false already signals this is a read/write tool. The description adds a concrete list of operations but doesn't disclose behavioral nuances such as side effects on the current drawing, file system changes, or reversibility beyond the operation names themselves. It's adequate but not deeply transparent.
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 efficiently structured as a bulleted list of operations, with each line providing a clear verb and optional data parameters. It's front-loaded with a one-line summary and avoids unnecessary prose.
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 many operations, the description covers each one's core purpose and data inputs, and includes special cases like QSAVE. It lacks details on prerequisites (e.g., whether a drawing must be open) and the incomplete `include_screenshot` flag, but overall it's reasonably 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?
The description goes beyond the schema by documenting the expected `data` structure for each operation (e.g., `{path}` for open, `{name?}` for create), which greatly helps parameter usage. However, it doesn't mention the `include_screenshot` parameter and uses an ad-hoc notation rather than aligning with the schema properties.
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 'Drawing file management' and enumerates specific file-level operations (create, open, save, export, plot, purge, undo, redo), distinguishing this tool from sibling tools that focus on layers, entities, blocks, etc.
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 groups operations under a single tool with a clear file-management scope, but it never explicitly states when to use this tool versus sibling tools like `layer` or `entity`. The usage is implied by the operation list, not directly stated.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Entity creation, querying, and modification.
Create operations: create_line — x1, y1, x2, y2, layer? create_circle — data: {cx, cy, radius}, layer? create_polyline — points: [[x,y],...], data: {closed?}, layer? create_rectangle — x1, y1, x2, y2, layer? create_arc — data: {cx, cy, radius, start_angle, end_angle}, layer? create_ellipse — data: {cx, cy, major_x, major_y, ratio}, layer? create_mtext — data: {x, y, width, text, height?}, layer? create_hatch — entity_id, data: {pattern?}
Read operations: list — layer? → list entities count — layer? → count entities get — entity_id → entity details
Modify operations: copy — entity_id, data: {dx, dy} move — entity_id, data: {dx, dy} rotate — entity_id, data: {cx, cy, angle} scale — entity_id, data: {cx, cy, factor} mirror — entity_id, x1, y1, x2, y2 offset — entity_id, data: {distance} array — entity_id, data: {rows, cols, row_dist, col_dist} fillet — data: {id1, id2, radius} chamfer — data: {id1, id2, dist1, dist2} erase — entity_id
| Name | Required | Description | Default |
|---|---|---|---|
| x1 | No | ||
| x2 | No | ||
| y1 | No | ||
| y2 | No | ||
| data | No | ||
| layer | No | ||
| points | No | ||
| entity_id | No | ||
| operation | Yes | ||
| include_screenshot | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description enumerates read and modify operations, and the annotations already signal readOnlyHint=false. It does not disclose side effects such as whether modifications are permanent, destructive actions like erase are irreversible, or how screenshot capture works. It adds the operation list but not deeper behavioral context.
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 efficiently structured into Create, Read, and Modify sections with each operation on a single line. It is front-loaded with a summary line and contains no filler. Every sentence earns its place in conveying the operation set and parameter usage.
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?
The description covers all operations and their parameters, which is comprehensive for a multi-operation tool. It omits the include_screenshot parameter and some operation-specific nuances (e.g., what fillet radius means), but given the output schema exists and annotations are present, it is sufficiently complete. The missing screenshot behavior is a minor gap.
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?
The schema provides zero descriptions for its 10 parameters, but the description maps each operation to its relevant parameters (e.g., create_line — x1, y1, x2, y2, layer?). This is valuable beyond the schema. However, it does not fully explain nested data objects (e.g., for create_hatch, data: {pattern?} is cryptic), so it earns a 4 rather than a 5.
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 opens with 'Entity creation, querying, and modification,' clearly stating the tool's purpose and scope. It then enumerates specific operations (create_line, list, erase, etc.), which distinguishes it from sibling tools like layer or drawing. This is a specific verb+resource statement with clear scope.
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 lists all sub-operations, implying when to use the tool for entity tasks. However, it does not explicitly state when to prefer this tool over siblings or provide exclusions (e.g., 'for layer management, use the layer tool'). The context is clear but lacks explicit guidance on alternatives.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Layer creation and management.
Operations: list — List all layers with properties. create — data: {name, color?, linetype?} set_current — data: {name} set_properties — data: {name, color?, linetype?, lineweight?} freeze — data: {name} thaw — data: {name} lock — data: {name} unlock — data: {name}
| Name | Required | Description | Default |
|---|---|---|---|
| data | No | ||
| operation | Yes | ||
| include_screenshot | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations only indicate readOnlyHint=false, which is consistent with the mutating operations listed. The description adds operation-specific data fields but does not disclose side effects, error behavior, or prerequisites. It covers the basic behavior but not deeper context like permissions or reversibility.
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 a brief intro and a clean bullet list of operations. Every line communicates meaningful information without redundancy or fluff. It is appropriately sized for the tool's complexity.
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?
The description covers all operations and their data parameters, which is essential for this multi-operation tool. Since an output schema exists, return value details are not required. However, it lacks any mention of error handling, side effects, or the purpose of include_screenshot, leaving minor gaps for a complete understanding.
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 description coverage is 0%, so the description must compensate. It does so thoroughly by defining the expected 'data' object for each operation (e.g., create: {name, color?, linetype?}, set_current: {name}). It also clarifies that 'operation' accepts the listed values, making the schema far more usable.
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 'Layer creation and management' and enumerates eight specific operations (list, create, set_current, etc.) with a verb and resource. This makes the tool's purpose unambiguous and distinguishes it from sibling tools like 'entity' or 'block'.
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 implies usage for layer operations but does not explicitly state when to use this tool vs alternatives. There are no exclusions or references to sibling tools. The operation list gives context, but the 'when-to-use' guidance is only implicit.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
P&ID drawing with CTO symbol library.
Operations: setup_layers — Create standard P&ID layers. insert_symbol — data: {category, symbol, x, y, scale?, rotation?} list_symbols — data: {category} draw_process_line — data: {x1, y1, x2, y2} connect_equipment — data: {x1, y1, x2, y2} add_flow_arrow — data: {x, y, rotation?} add_equipment_tag — data: {x, y, tag, description?} add_line_number — data: {x, y, line_num, spec} insert_valve — data: {x, y, valve_type, rotation?, attributes?} insert_instrument — data: {x, y, instrument_type, rotation?, tag_id?, range_value?} insert_pump — data: {x, y, pump_type, rotation?, attributes?} insert_tank — data: {x, y, tank_type, scale?, attributes?}
| Name | Required | Description | Default |
|---|---|---|---|
| data | No | ||
| operation | Yes | ||
| include_screenshot | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description lists mutating operations (setup_layers, insert_symbol, etc.), consistent with readOnlyHint=false. However, it does not disclose side effects, persistence requirements, or whether operations are reversible. It provides more than a tautology but lacks deeper behavioral context like permission needs or transaction behavior.
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 efficiently structured as a bullet-like list with no redundant prose. It front-loads the purpose and then lists operations with their data fields in a scannable format. A few operation details are terse, but overall the compactness is appropriate for the information conveyed.
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 existence of an output schema (not shown) and generic input schema, the description provides the essential operation dictionary and data shapes, which is critical for correct invocation. It lacks examples or enumerations of valid symbol/category values, but for a tool of this complexity, it covers the main invocation paths adequately.
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 0% for the top-level parameters, but the description compensates by specifying the allowed 'operation' values and the expected 'data' payload for each operation (e.g., insert_symbol with category, symbol, x, y). It does not explain the 'include_screenshot' parameter, which is a minor gap. The detailed operation list provides substantial semantic value beyond the generic 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 identifies the tool as 'P&ID drawing with CTO symbol library' and enumerates specific operations (setup_layers, insert_symbol, etc.), making the purpose evident. It does not explicitly contrast with sibling tools, but the focused domain and operation list distinguish it from generic tools like 'entity' or 'drawing'.
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 implies when to use the tool (for P&ID drawing operations) but does not explicitly state alternatives or exclusions. It offers no guidance such as 'use this for P&ID symbols, use entity for generic entities', but the operation list itself conveys that this is the place for P&ID-specific tasks.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Server status and management.
Operations: status — Backend info, capabilities, health check. health — Quick health check (ping backend). get_backend — Return current backend name and capabilities. runtime — Return process/runtime details for spawn diagnostics. init — Re-initialize the backend. execute_lisp — Execute arbitrary AutoLISP code (File IPC only). data: {code}
| Name | Required | Description | Default |
|---|---|---|---|
| data | No | ||
| operation | Yes | ||
| include_screenshot | No |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations declare readOnlyHint=true, yet the description includes 'init — Re-initialize the backend' and 'execute_lisp — Execute arbitrary AutoLISP code', both of which are mutating/executing operations. This is a direct contradiction between structured annotation and description.
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 summary line is followed by a compact bulleted list of operations; every line adds specific information. No filler or 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?
The description covers all six operations and their purposes, which is strong for a multi-tool. However, it omits the include_screenshot parameter and provides no guidance on when to use this tool vs siblings; the contradiction with readOnlyHint further undermines 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?
Schema coverage is 0%, and the description compensates by enumerating valid operation values and noting data:{code} for execute_lisp. However, include_screenshot is never explained, and the data parameter is only partially specified.
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?
Description opens with 'Server status and management' and enumerates six specific operations (status, health, get_backend, runtime, init, execute_lisp). This clearly identifies the resource (server/backend) and distinguishes it from sibling tools focused on drawing/entities.
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 when-to-use or alternative guidance is provided. The operation list implies usage contexts (e.g., health checks, runtime diagnostics), but the description never states when to choose system over sibling tools or which operations are appropriate for which scenarios.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Viewport control and screenshot capture.
Operations: zoom_extents — Zoom to show all entities. zoom_window — Zoom to window: x1, y1, x2, y2 get_screenshot — Capture current view as PNG image.
| Name | Required | Description | Default |
|---|---|---|---|
| x1 | No | ||
| x2 | No | ||
| y1 | No | ||
| y2 | No | ||
| operation | Yes |
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With readOnlyHint=true in annotations, the safety profile is already declared. The description adds behavioral context by enumerating the specific operations and noting that get_screenshot captures the current view as a PNG image. It does not contradict the annotation, and the mention of coordinate parameters for zoom_window provides some operational transparency beyond the raw schema.
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 one-line summary followed by an operation list. It front-loads the main purpose and uses a bullet format for operations, making it easy to scan. Every sentence adds value, with no redundant or extraneous content.
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 moderate complexity (5 params, 1 required, no enums) and the presence of an output schema, the description covers the main operations and their associated parameters. It does not detail edge cases like invalid operations or coordinate requirements, but the essentials are present. The output schema presumably handles return-value documentation, so the description is mostly 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?
The schema has 0% description coverage, leaving the description to compensate. The description does add meaning by linking x1, y1, x2, y2 to the zoom_window operation, and it names the operations. However, it does not explain the coordinate system, units, or whether all coordinates are required for zoom_window, so parameter semantics remain partially under-specified.
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 identifies the tool as controlling the viewport and capturing screenshots, with specific operations listed: zoom_extents, zoom_window, and get_screenshot. This is a specific verb+resource combination that distinguishes it from sibling tools like layer or entity, which handle different aspects.
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 implies usage by listing operations and their parameters (e.g., 'zoom_window — Zoom to window: x1, y1, x2, y2'), but it does not explicitly state when to use this tool versus alternatives or provide exclusions. For a viewport-focused tool, the usage context is fairly clear from the operation list, but without explicit alternative guidance it remains at the 'implied usage' level.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
The eight tools cover distinct domains (layers, drawing, entities, blocks, annotations, PID, view, system) with clear purposes. However, entity.create_mtext and annotation.create_text both handle text creation, which could cause confusion about which tool to use for text.
All tool names follow a consistent single-noun pattern (layer, drawing, entity, block, etc.), and within each tool, operations consistently use verb_noun style (create_line, set_current, etc.). The naming convention is uniform and predictable.
With 8 tools, each representing a major functional area of AutoCAD, the count is well-scoped and avoids unnecessary granularity. The breadth covered (drawing management, entity creation, annotation, P&ID, system) is substantial but each tool earns its place.
The tool set covers most core workflows: drawing lifecycle, entity CRUD, block management, annotations, and system maintenance. Minor gaps exist such as no layer deletion/renaming and no modification operations for annotations, but these are not critical for typical automation tasks.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
MCP server for generating rough-draft project plans from natural-language prompts.
Nifty's MCP server — exposes tasks, projects, messages, and files as tools for AI agents.
MCP server for progressive tool usage at any scale (see https://klavis.ai)
MCP server unifying ERPs, CRMs, APIs and knowledge base for Claude, ChatGPT and Gemini.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/gpt2nndk/autocad-mcp-nndk'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
