An ISA-5.1 signal line has no pipe line number, yet every pneumatic /
electric line got the bare sequence as its number and a label, and ate a
process sequence number. Signal classes now leave line_number None (a
verbatim one is still honoured), seq None, no label. A radial port aimed
at the far end even with waypoints, so the first segment cut through the
bubble; aim_points (lines.py) aims at the first/last waypoint for both
draw_line and the dry run.
Merge note: benchmarks/adapters/autocad_mcp_pro.py (final group's file)
had its two _task_pid_roundtrip assertion lines moved to the new truth —
the signal row is unnumbered, the connector row is 3.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Two review findings on 32821c6.
1. Once the shielded rollback body runs after a real client cancellation it
can fail (the COM hard timeout, _safe_send_command's own deadline, an
ezdxf quarantine refusal). That RuntimeError escaped run_spec as an
ordinary error and the mcp lowlevel server, which had already answered
the cancelled request, tried to respond a second time: `assert not
self._completed` in mcp/shared/session.py escaped the server task group
and the whole session died (measured through the real transport: the next
system_status was never answered and Client.__aexit__ re-raised the
AssertionError). run_spec now decides by the exception it is handling:
after a cancellation the rollback error is logged and the cancellation is
what propagates, so the SDK's cancellation suppression applies; after any
other error the rollback error propagates as before, with the original
chained as its context.
2. The docstring promised the rollback on both engines, but on the live
AutoCAD 2026 ComBackend.transaction_rollback sent `_UNDO B` with no UNDO
Mark in the stack (StartUndoMark/EndUndoMark form an undo *group*, not a
Mark), AutoCAD answered "This will undo everything. OK? <Y>", SendCommand
blocked at that prompt, _run gave up after 60 s, the entities stayed and
every COM client was rejected until a human pressed ESC. transaction_begin
now types `_.UNDO _M` before StartUndoMark (refused with ok:false while
CMDACTIVE, where the Mark would be swallowed as input), and
transaction_rollback ends the group and sends `_.UNDO _B`, which returns
exactly to that Mark - also on an empty transaction, where `_UNDO 1` would
be one step too many. A rollback with no transaction is refused the way the
headless engine refuses it, because an unmarked Back is that prompt.
Measured with the real ComBackend in a scratch Documents.Add() document on
the running AutoCAD 2026 (closed with Close(False)): begin 31 ms ->
entity_create_line + circle -> rollback 63 ms -> 3 entities back to 1,
CMDACTIVE 0, no prompt; an empty begin/rollback leaves the older line;
begin/commit unchanged.
Tests: the two failure modes on run_spec (anyio scope cancellation with a
raising rollback completes the task group and logs; a ValueError with a
raising rollback still raises the rollback error with the ValueError as
context), the real-transport case (client.cancel mid-run, rollback raises,
the next request is answered, cleanup through the real rollback), the COM
shape now pins `_.UNDO _M, StartUndoMark, EndUndoMark, _.UNDO _B`, and
test_com_sprint3 grows a Mark-keeping fake document whose unmarked Back
fails the way the prompt does: begin before the group, rollback consumes the
Mark, rollback without a transaction sends nothing, begin refuses on
CMDACTIVE.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
_num let NaN/inf (and a 400-digit int, as OverflowError) through, so an
INSERT was placed at [nan, 0] and the refusal, when it came, named the
line instead of equipment[0].x. run_spec caught Exception only, so a
CancelledError mid-run left the placed symbols inside an open transaction
(an open undo mark on COM). Finite check in _num; except BaseException.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
base_x/base_y were only float()-cast, so a NaN base point was written into
the BLOCK record; a primitive on a layer the drawing lacks was written as a
dangling reference doc.audit() never flags (and on COM would fail mid-loop
after Blocks.Add). validate_base_point names the key; the layer table is
probed before the first write and the missing names are refused, or created
and reported as layers_created when create_layers=true (new optional kwarg,
contract + both engines + tool).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
block_insert and block_set_attributes applied str() to whatever arrived, so
a JSON null became the text None and a non-mapping failed after the INSERT
was written. validate_attribute_values (backends/block_specs.py) runs first
on both engines: strings pass the one-line rule, ints and finite floats are
written as plain digits, everything else is a TypeError naming the tag.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Headless text_find_replace matched and wrote only dxf.text of an ATTRIB/ATTDEF, so on a multi-line attribute it reported replaced: 1 while block_get_attributes and block_explode still carried the old embedded MTEXT (and a needle on the second line was never found). The ATTRIB/ATTDEF branch now reads through _attrib_value and writes through _write_attrib_value, the same pair block_get_attributes / block_set_attributes use.
COM block_explode swallowed a failing GetAttributes() (attrs = ()), GetConstantAttributes() and Invisible, then exploded, deleted every ATTDEF placeholder and the reference and returned ok: True with no attribute text -- every value destroyed silently on a transient RPC_E_CALL_REJECTED. Those reads now happen before Explode() and propagate with nothing written; a failure after the explode deletes the members it added and the TEXTs written so far and leaves the reference in place (_undo_explode), so the drawing is never left half-burst under an exception. The fakes gain the members a live AcadBlockReference / AcadAttributeDefinition always expose (GetConstantAttributes, Constant, Invisible) instead of relying on the swallow.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Two review findings on the headless engine, both content lost with ok: True.
Multi-line attribute writer: `block_set_attributes` set only `attrib.dxf.text`,
so after an edit `block_get_attributes` reported the new value while
`block_explode` (`_burst_attrib`) burst the embedded MTEXT at its pre-edit
content -- and on the default R2010 document, where ezdxf exports no embedded
MTEXT, a save/reload carried the new value the exploded drawing never showed.
`_write_attrib_value` now rebuilds the embedded MTEXT from its own virtual
entity with the new content (`virtual_mtext_entity()` + `set_mtext(...,
graphic_properties=False)`, the round trip ezdxf's `BaseAttrib.transform`
uses), keeping placement, style and height; `_attrib_value` reports the
embedded MTEXT's full `Line1\PLine2` content for a multi-line attribute, as
AutoCAD's `TextString` does, so get / set / explode / COM agree.
Opaque members: `block_explode` iterated `virtual_entities()` without a
`skipped_entity_callback`, so ezdxf dropped every member it cannot copy
(OLE2FRAME) or transform (VIEWPORT) with a debug log, an ACAD_PROXY_ENTITY
without proxy graphics contributed nothing, and the reference was then deleted.
`_explodable_members` materializes the list before any write and refuses by
member with the typed capability `explode_opaque_members` (declared on both
engines: headless False with the reason, COM native), naming the block and
the live way out. The block_explode docstrings (backend and tool) now state
this refusal instead of claiming the xref/MINSERT gate covered everything.
Tests: eight new headless cases in tests/test_track_a_hardening.py -- the
multi-line reader, the rewrite (value on every surface, placement unchanged,
clean audit), a `\P` value and the empty value, an R2010 save/reload, the
single-line path unchanged, the opaque-member refusal (nothing written,
reference alive, members named), an ordinary block still exploding under a
uniform scale, and the capability declared on both engines.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
A multi-line ATTRIB keeps its content in an embedded MTEXT; dxf.text is only
the first line (ezdxf) or "Line1\PLine2" (AutoCAD), so the TEXT the explode
emitted silently dropped lines or rendered literal \P with a clean audit.
ezdxf: `_burst_attrib` adds `virtual_mtext_entity()` (graphic properties and
WCS placement already carried; the reference matrix applied for a constant
ATTDEF) and marks it invisible like the ATTRIB. COM: `_capture_attrib` reads
`MTextAttribute` / `MTextAttributeContent` / `MTextBoundaryWidth` and
`_add_mtext_from_capture` goes through `AddMText` (height, style, frame,
then Rotation, then the attachment mapped from the ATTRIB's Alignment).
Two INSERT kinds `virtual_entities()` cannot explode were destroyed with
ok: True: an xref (no virtual entities, reference deleted) and a MINSERT
(only the first cell resolved, the rest of the grid gone). Both are now
refused by name before any write, on both engines (`block_record.is_xref`
/ `mcount > 1` headless; `Blocks.Item(Name).IsXRef` / `AcDbMInsertBlock`
live, the latter previously failing the INSERT check with the misleading
"not a block reference").
Tests: nine new cases in tests/test_track_a_hardening.py (headless multi-line
ATTRIB and constant ATTDEF, xref and MINSERT refusals, a single-cell grid
that still explodes; COM multi-line capture, frame-before-angle, and both
named refusals). The fake ActiveX surface gains AddMText and IsXRef.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
COM: ``text.Normal`` was a plain Python tuple, which ActiveX marshals as
VT_ARRAY|VT_VARIANT and refuses with E_INVALIDARG (verified live on
AutoCAD 2026); because the write came after Explode(), the ATTDEF deletes
and AddText, the tool raised and left the drawing torn. It now goes
through ``_apoint`` like every other point write, and the fake ActiveX
surface refuses a bare tuple on Normal/InsertionPoint/TextAlignmentPoint
so a test cannot pin the wrong call shape again.
COM: every AcDbAttributeDefinition Explode() returned was deleted, but
GetAttributes() excludes constant attributes, so a constant attribute's
only visible text was destroyed with ok: True. A constant ATTDEF from the
explode (Constant member, or its tag in GetConstantAttributes()) is now
captured like an ATTRIB and replaced by a TEXT; placeholders are still
just deleted.
ezdxf: an AutoCAD-authored reference carries no ATTRIB for a constant
attribute and virtual_entities() skips ATTDEF, so the same text vanished
headlessly. A constant ATTDEF whose tag has no attached ATTRIB becomes a
TEXT transformed by the reference matrix, the path add_auto_attribs takes,
and is not converted twice when block_insert did attach one.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Two silent wrong-place defects in the BURST path. An ATTRIB is an OCS
entity, so the TEXT that replaces it now carries extrusion/thickness with
insert/align_point; a mirrored reference (ATTRIB reflected to (0,0,-1),
INSERT +Z with yscale=-1) put the tag 50 units across the mirror axis with
a clean audit. Members and TEXTs now go into the INSERT's owner layout
(ezdxf get_layout, COM ObjectIdToObject(OwnerID)) instead of the current
space: a sheet title block exploded while Model was current duplicated
geometry into model space and then died on delete_entity headless, and on
COM landed the tag text in model space with ok: True. A reference nested
in a block definition is refused before any write on both engines.
Five regression tests (mirrored INSERT, sheet-owned INSERT with Model
current, nested INSERT, and the COM owner-block / nested-refusal call
shapes) fail on the previous commit and pass here.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
virtual_entities() yields geometry only, so an exploded tagged symbol lost
its tag value silently. Every attached ATTRIB now becomes a TEXT at the
same WCS placement/height/rotation/style/layer (invisible stays
invisible), reported as attribute_texts. COM reads GetAttributes before
Explode, deletes the ATTDEF placeholders Explode returns, AddText per
value, and deletes the reference ActiveX Explode leaves behind.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Track A's task-1 gate covered block_insert only; the sibling tool still
returned a handle for a typo'd name that doc.audit() later deleted, and
both tools accepted *Model_Space (a reference cycle). One gate per engine
(_require_block / _require_block_defined) is now shared by both placing
tools: non-string name -> TypeError, unknown -> ValueError naming it,
layout block -> ValueError, all before add_blockref / InsertBlock.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
TOOL_PACKS=all (default) | comma list of core,pid. core is everything not
claimed by another pack and can never be dropped; unknown entries are
ignored with a warning. Pack hiding composes with TOOL_PROFILE=lean (the
lean set now carries pid_symbol_insert / pid_line_draw / pid_graph, 50
tools) and ENABLE_3D. system_about reports {enabled, available, ignored};
cad_batch's unknown_tool rejection names the TOOL_PACKS gate.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Closes the four verified BLOCKING findings on pid_graph.
1. scope="all" attached across layouts: the port grid, bbox candidates,
vertex grid, junctions and text lookup were pooled from every space.
The grid is now keyed by (space, cell) and every lookup filters by the
edge's space; junctions carry `space`.
2. A mirrored INSERT (extrusion -Z, as MIRROR3D / foreign DXFs store it)
got its catalogue ports on the unmirrored image at 1.0. Both engines
report `mirrored` on INSERT properties (ezdxf: extrusion z < 0; COM:
Normal z < 0; default False); transform_port takes `mirrored` and
applies x -> -x about the WCS insertion point after rotation/scale,
direction (180 - d) mod 360 — cross-checked against Insert.matrix44.
resolve_endpoint passes it through too, so draw_line starts on the
mirror image as well.
3. The text-tag fallback searched a full circle around the box top, so a
line label or note below the symbol beat the tag above it. Candidates
must now sit at or above top - height/2 (spec §9.2 step 6), nearest
wins, and line-number texts are never equipment tags.
4. A non-uniform INSERT scale re-ran the transform with the X factor and
set `inferred` (reserved for foreign-block ports). transform_port
gains `strict=False` for the reader: point ports scale per axis
exactly; a radial port keeps min(sx, sy)·r and the node drops to 0.6
with notes ["non-uniform scale"]. Node dicts gain `y_scale`,
`mirrored` and `notes` (default []).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FKFStjEFzpGyLq5bMpWxTd
Every one of these was reported ok/clean to the caller while being wrong.
Silently wrong numbers
- selection_filter(min_area) read `entity.dxf.get("area", 0.0)` for SOLID,
TRACE and HATCH, none of which carry that attribute, so every hatch
measured 0.0. A 1600 mm^2 hatch was excluded by min_area=1.0 while
`filtered_by` asserted the filter had run, and COM (which reads ActiveX
.Area) returned a different selection for the same call. It now answers
from the measurement `entity_measure` uses, so the two cannot disagree.
- Every headless dimension rendered through ezdxf's bare `Standard` style:
1.0 mm text with a comma decimal marker, against ISO 3098's 2.5 mm minimum
and ISO 129's point -- while the document header itself said $DIMTXT 2.5.
`ezdxf.new(setup=True)` is not the fix; measured, it renders a 100 mm
dimension as "10000". The attributes are now stored explicitly.
- Saving writes `dimrnd = 0.0` onto the Standard style, and ezdxf applies
`xround(value, dimrnd)` whenever the attribute is present at all --
`xround(33.333, 0.0)` is 33. So on any reopened file, including one this
server saved, every dimension was rounded to a whole unit before DIMDEC
formatted it: a 12.75 mm feature dimensioned 13. AutoCAD means DIMRND 0 as
*no* rounding; the sentinel is now discarded at every load site, and a real
0.5 mm rounding is left alone.
- The DIM* header fold gave out-of-range values teeth they did not have when
they died in the header: dim_decimals=-1 returned ok and then killed the
next dimension with `ValueError: Format specifier missing precision` from
inside ezdxf, and 40 printed forty digits of float noise onto the sheet.
Ranges are now AutoCAD's own (DIMDEC 0-8, DIMZIN 0-15, lengths positive).
Security
- security.py recorded the policy as "sanitize_command and sanitize_lisp each
have exactly ONE caller ... the policy is scoped to the CHANNEL". The
premise was false. `_ensure_linetype_loaded` and `ComBackend.linetype_load`
build `_-LINETYPE _LOAD {name} {file}` and hand it to SendCommand, which
treats a newline and ';' as segment separators -- so a `linetype=` argument
carrying either injected further AutoCAD commands, and the 36-verb denylist
never saw that channel. Guarded with the DXF symbol-name rule (the
injection characters fall out of it rather than being enumerated) plus
validate_path for the file argument, which was also the one path-taking
tool reaching the filesystem unchecked. The policy comment is corrected.
Gates that could not see
- drawing_critique returned zero issues on a quarantined document: all seven
checks wrapped their backend call in a bare `except Exception`. CLAUDE.md
rule 7 makes that call the gate an agent is told to trust. Only the
quarantine now escapes; an ordinary broken check still degrades.
`_check_dimension_legibility` also read `_doc` directly from the event loop
thread, missing the gate and performing exactly the unsynchronised access
the quarantine exists to prevent. `construction_clear` answered
"deleted: 0" -- there was nothing to clear -- for the same reason.
- The iso128 focus checked layer lineweights only, so a sheet lettered at
1.0 mm critiqued clean. It now reads the effective dimension height and
separator, header-then-style, in the order the renderer resolves.
- check_doc_numbers.py derived five values it never checked, and those were
the ones that had drifted. The collected-test count (README said 1274, the
tree had 1369) and the A/B pass-rate and Fixed columns are now gated.
- benchmarks/README.md's v1.4.0 table was impossible on its face -- 21
passing plus 5 fixed is 26, not 24 -- and contradicted the report it
linked, because `ab-v1.4.0-vs-v1.5.0.json` was regenerated from a tree that
already carried the dimension fixes and so never measured the v1.5.0 tag.
Replaced with a real run: ab-v1.4.0-vs-v1.5.1.json, 21/26 -> 26/26.
Release gate
- release.yml ran one ubuntu/3.11 lane, so a Windows-only defect could reach
PyPI with no Windows job having executed. That is not hypothetical: v1.5.0
shipped `except pywintypes.com_error` raising NameError wherever pywin32 is
absent. It now runs the same three lanes as ci.yml, Windows included.
1368 passed, 1 xfailed. ruff clean. Tool inventory and all 16 gated document
numbers match the tree.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Three defects found after 1.5.0 reached PyPI, two of them by the adversarial
pass over the rewritten README.
The HTTP bind guard failed open on the launch path it was written for.
`_guarded_run_async` read the bind address from the `host` kwarg and treated an
absent one as loopback. It is not: `fastmcp run` forwards that kwarg only when
`--host` was passed, and fastmcp otherwise resolves the host from
`fastmcp.settings.host`, which is environment-backed. So
FASTMCP_HOST=0.0.0.0 fastmcp run server.py:mcp --transport http
bound every interface -- 149 tools including file open/save and AutoCAD command
execution, no token -- while the guard inspected the literal "127.0.0.1" and
waved it through. v1.4.0's changelog claimed this guard "fires on every launch
path"; it fired, and looked at the wrong host. It now resolves the effective
host, and warns rather than staying silent if it cannot.
`[com]` and `[full]` could not be installed off Windows. Both listed
`pywin32>=306` with no environment marker, and pywin32 publishes Windows wheels
and no sdist, so `pip install autocad-mcp-pro[full]` died at resolution on
Linux and macOS -- and `[full]` is the extra the README recommended to
contributors. CI never caught it because every Linux lane installs `[pdf]` or
syncs from uv.lock, whose resolution markers hide the gap. Both entries carry
`; sys_platform == "win32"` now, with a release-consistency test to keep them.
`drawing_settings({"dimscale": ...})` reported success and changed nothing
headlessly. It writes `$DIMSCALE`, which is what AutoCAD reads when it draws a
dimension, so the call worked on COM; ezdxf renders from the dimstyle table
entry and never consults the header. One call, one reported success, two
different drawings. The header value now folds into the per-dimension override
on all five dimension tools -- including `dimension_aligned` and
`dimension_angular`, which do not share the tolerance-override path and the
obvious fix would have missed. DIMSCALE 1.0 writes nothing.
README: every engine-conditional family in the feature table now says which
engine (boundary tracing, WIPEOUT/REVCLOUD/MTEXT masks, CHSPACE, typed hatch
edge paths, headless undo), and the checkout install line stopped recommending
an extra that could not resolve on the reader's platform.
1257 passed, 1 xfailed. 1.5.0 is to be yanked: its diameter and radius callouts
put wrong numbers on drawings at default settings.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`add_diameter_dim` / `add_radius_dim` measure to the point they are handed;
`leader_length` is a text placement. The ezdxf backend passed
`centre + (radius + leader)` as that point, so the leader became part of the
measurement: a true 40 mm bore was dimensioned 60 at the default
`leader_length=10`, and 108 at 34. Radius callouts were `leader_length` too
large the same way.
This is the worst class of defect for this project -- a wrong number, at
default settings, on the one part of a drawing that exists to carry numbers --
and it shipped in v1.4.0 and earlier. The size now comes from `radius` and
`location` only places the text.
The live COM backend was never affected: ActiveX's AddDimDiametric takes the
two chord points and the leader length as separate arguments. The two engines
had been disagreeing by 2 x leader_length on the same call.
Found while building the drawing for the new README, whose own bore callout was
wrong. Two checks added to correctness_suite.py; both report fail -> pass
against v1.4.0, not miss -> pass, which is what separates a repaired defect
from new capability in that lane. A/B is now 21/26 -> 26/26, five fixed, zero
regressed.
README rewritten from scratch
- Hero is a real ISO A3 sheet the server drew: m6 z24 involute spur gear, DIN
6885 keyed bore, section A-A, ISO 129 dimensions, ISO 286 H7 fit, ISO 7200
title block, 56 entities, drawing_critique clean. New
scripts/render_readme_showcase.py rebuilds it, and the tool-call sequence
that produces it is printed in the page.
- Star history chart, badge row, capability matrix, five evidence lanes, and a
"what this release is bad at" section carrying the 28% throughput regression.
- Every image and repo link is now an absolute URL: README.md is the PyPI long
description, where relative paths 404.
Corrections this work forced, all measured
- The df=0 example was wrong. FILLET is already in entity_fillet's own text;
the commands that genuinely scored zero are WBLOCK, PEDIT, OVERKILL,
LAYTRANS and MATCHPROP.
- 40,305 tokens is the tools/list wire payload; ~25,143 is the part forwarded
to the model. The old wording said "of schema", which is neither.
- `--tokenizer anthropic` needs a package and credentials this repo does not
lock, and refuses rather than falling back. Said so.
- Discovery ranks over the 149-tool advertised catalog, not the 154 registered.
- Five middlewares, not four; the fifth is what makes refusals machine-readable.
- Five of the 27 capability keys are environment-dependent, so the matrix
carries footnotes instead of bare ticks.
- "Every coordinate is WCS" has one deliberate exception: TEXT rotation.
- render_live_chart's axis still said "fixed 10-task matrix" while our bar came
from v3's fifteen. Each bar now carries its own matrix.
- .env.example said MAX_UNDO_STACK=50 (code: 5) and omitted four real
variables, two of them the remote-HTTP controls.
- server.json described an empty ALLOWED_PATHS as blocking access. It disables
the allowlist.
- server.py called the lean profile ~46 tools; it is 47.
- backends/capability.py cited a test file that does not exist.
1249 passed, 1 xfailed.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Tool count 131 -> 154; 474 -> 1242 collected tests.
The release exists for two measured reasons. A 149-tool catalog costs a client
~40,300 tokens before it has asked for anything, and a drafter searching the way
drafters think -- FILLET, BPOLY, QSELECT, WBLOCK -- scored df=0 against the stock
index, because those words were simply not in the data.
Discovery and batching
- DISCOVERY_MODE=search replaces the catalog with search_tools + call_tool,
ranked over an AutoCAD command and synonym corpus (discovery/aliases.py) on
top of fastmcp's BM25 index. Measured: 40,305 -> 356 tokens advertised.
- cad_batch runs a step list in one round trip; fields= projects result-heavy
responses. The `core` profile is gone -- discovery replaced it.
Honesty debt (Tier 0)
- OCS/ECS: `extrusion` appeared nowhere in this codebase, so mirrored and
rotated geometry reported coordinates that were quietly wrong. backends/ocs.py
translates at the backend boundary; every tool coordinate is WCS on both
engines, and a tilted plane refuses rather than answering with a number that
looks fine.
- drawing_save_as(fmt="dwg") wrote DXF bytes to a .dwg path. drawing_audit never
repaired. analysis_measure_area shoelaced points the caller typed in.
- ezdxf calls gained a per-call timeout, so one hung call can no longer wedge a
server whose document lock is a single asyncio.Lock.
Architecture
- The 1854-line backend ABC is 19 contract modules; @capability(key, reason=)
makes optional methods declarative and a gate fails on undeclared holes.
- 8 layout/viewport tools, CHSPACE, CAD_PROGID escape hatch.
- Wave A: selection filters, boundary tracing, hatch depth, annotation objects,
measurement -- 13 tools of a planned 39. The REGION and 2D-boolean family was
cut on evidence: add_region() yields zero ACIS bytes and the greiner_hormann
substitute loses 28.2% of the area on a square with one semicircular edge.
Verified against a live AutoCAD 2026, which found four COM defects no amount of
reading would have caught -- AcadPViewport has no ViewCenter, and the line that
set it had been swallowed by a bare except since v1.4.
Benchmark v3 (15 tasks) and what it caught
- v2 scored this server 10/10, which carried no information: every task in it
exercised something the server was built around. Five tasks were added because
they can fail, and three did while being written -- a hatch could not be
measured though the refusal listed HATCH as measurable, the boundary tools and
the measure tool disagreed about the same shape (139.2177 vs 139.2699), and a
circle stored as two bulged vertices measured 0.0.
- Competitor reports are NOT re-scored. They are pinned v1.4 runs against v2 and
were never asked these five questions; the chart shows those rows `not run`.
- Correctness A/B: 21/24 -> 24/24, zero regressions.
Performance, stated plainly
The published v1.4 perf report was recorded on CPython 3.14 and the new one on
3.11.15, so the apparent 2-6x speedup is the interpreter, not this release.
Re-measured on one interpreter, v1.5.0 is 28% slower creating entities and 22%
slower on the 10k roundtrip. Fully attributed: EZDXF_CALL_TIMEOUT=0 restores
v1.4.0's exact numbers. That is the deadlock fix being paid for, the knob is
documented, and winning it back is on the 1.6 roadmap.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
- TOOL_PROFILE=lean|core|full applied in the lifespan via server-level enable/disable; system_about reports the active profile; opt-in solid_* tools are hidden while ENABLE_3D=false
- layouts: layout_list/create/set_current + viewport_create (scaled model viewports) on both backends; drawing_export_pdf(layout=...); ezdxf uses TILEMODE + active-layout binding, COM switches ActiveLayout and restores it
- ISO 286: engineering/fits.py authored IT4-11 + fundamental-deviation tables (1-500 mm; shafts d..p/h/js, holes D..H/JS) with clear out-of-scope errors; dimension_linear/radius/diameter gained fit= resolved from the measured nominal
- tolerances: DIMTM now signed in deviation/limit modes so double-positive fits (p6 +0.035/+0.022) render correctly; legacy positive-magnitude contract preserved
- solids: solid_box/cylinder/extrude/revolve/boolean native on COM (AddBox/AddCylinder/AddRegion+AddExtrudedSolid/AddRevolvedSolid/Boolean); ezdxf reports the honest solid_3d capability boundary
- capability maps: paper_space native, new viewport_render feature, ENABLE_3D-gated solid_3d
- 60 new tests (profiles, layouts, viewports, layout PDF, fits, gating); version pins in tests now read version.__version__
Co-authored-by: Cursor <cursoragent@cursor.com>
Adds production-grade engineering features and tools: 2D GD&T (feature control frames + datum features) via engineering/gdt.py, ISO‑129 dimension tolerance support (engineering/tolerances.py) and an ezdxf/COM plumbing for toleranced dims, a scalar drawing score + invalidity ratio (engineering/scoring.py) wired into drawing_finalize, in-place entity editors (entity_edit_text/entity_edit_geometry), and a friendly drawing_settings facade. Includes server tool wiring, COM/ezdxf backend implementations, datum-consistency critique, many unit tests, README/CLAUDE/CHANGELOG updates, and a version bump to 1.2.0. Also corrects COM backend capability label (entities_2d).
New COM-only Entity Query tool that reads AutoCAD's implied ("pickfirst")
selection set and returns the selected entity handles + EntityInfo, so the AI
can scope work (e.g. dimension_auto) to exactly what the user picked in the
viewport instead of the whole drawing. Resolves#1 — the layer-juggling
workaround is no longer needed.
- backends/base.py: abstract selection_get() returning a status dict
- backends/com_backend.py: reads ActiveDocument.PickfirstSelectionSet, surfaces
the PICKFIRST sysvar, handles the empty-selection case
- backends/ezdxf_backend.py: headless stub (ok=False, empty handles, same shape)
- server.py: @mcp.tool registration (readOnlyHint), SECTION 5 3 -> 4 tools
- README + CHANGELOG + pyproject: 110 -> 111 tools, v1.1.0 -> v1.1.1
Verified live on AutoCAD 2027 (COM): selection_get correctly returned the
user's grip selection across two distinct picks; 318 tests pass, ruff-clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Audit-driven fixes across the dual backend, the premium workflow, and the
engineering layer. 318 tests, ruff-clean.
Quality gate (the moat):
- drawing_finalize now runs the premium ISO-128 critique in addition to the
8-step validator; leftover construction, non-ISO lineweights, untrimmed
corners, duplicates and dim overlap block the gate (strict_critique fails on
warnings too).
- dim_overlap critique no longer a no-op (dim reference point + bbox fallback);
dim/construction layers resolve from the active layer set (iso13567 too).
Stability / correctness:
- view_screenshot & drawing_export_pdf render headless via Agg (no more
worker-thread SIGSEGV).
- gear tooth profile no longer self-overlaps at high tooth counts; section view
no longer duplicates bore lines; validator keyway heuristic fixed.
- ezdxf aligned/angular dim API, full-circle polar array duplicate already
covered by Sprint 1; this bundle adds save-format-from-extension,
entity_offset side selection, ARC length, properties parity (bounding_box /
rotation / char_height), mtext rotation.
COM robustness (mocked-COM tested):
- CoUninitialize on teardown; transaction commit/rollback try/finally + UNDO via
the CMDACTIVE guard; system_run_lisp guard + USERS1 value capture;
set_variable type coercion; HWND via Application.HWND; BlockInfo.description.
Security:
- HTTP remote-bind guard fires on every launch path (incl. `fastmcp run`);
StaticTokenVerifier wired into FastMCP; AutoLISP allowlist bypass-vector tests.
Hygiene: linetype-on-demand in _apply_attrs, honest view_zoom returns, snap-on-arc
sweep handling, point_tangent acos clamp, dynamic tool_groups / tool count,
word-boundary SPUR/HELICAL, undo-stack locking.
Deferred: R22 (undo/transaction stack isolation) lands with the I1 refiner (1.2),
which depends on it.
Before: _validate_http_bind warned about MCP_AUTH_TOKEN being required for
remote HTTP but the token was never passed to the FastMCP constructor, so the
server accepted any request without authentication.
After: a StaticTokenVerifier is built at module load if MCP_AUTH_TOKEN is set
and passed as auth= to FastMCP(). Loopback-only deployments (no token set)
continue to work as anonymous mode.
7 new tests in tests/test_http_auth.py covering: token present/absent, mcp
instance wiring, loopback/remote bind guard, remote-without-token rejection.
233 tests total.
New concrete base methods (backend-agnostic, use entity_get + pure math):
- AutoCADBackend.point_intersection: LINE-LINE, LINE-CIRCLE, CIRCLE-CIRCLE
with ref_x/ref_y disambiguation for multiple candidates.
- AutoCADBackend.point_tangent: external-point-to-circle tangent touch-point
with ref_x/ref_y for side selection; correct perpendicularity formula
(T-F)⊥(T-C) via cos(θ-α)=-r/d.
Registered as MCP tools: point_intersection, point_tangent.
13 new tests in tests/test_meta_tools_geometry.py.
226 tests total.
Introduce the autocad-mcp-premium skill with workflow docs, examples, and references (.claude/skills/autocad-mcp-premium/*). Add PlanSpec/critique plumbing and engineering helpers (engineering/plan_spec.py, engineering/critique.py) plus tests and a smoke script. Extend backend API (backends/base.py) with corner operations and premium meta-tools (trim/extend/fillet/chamfer, drawing_plan, drawing_critique, point_from_snap, construction_xline/clear, drawing_apply_iso_layers, dimension_auto, entity_select_smart) and implement COM backend corner ops (backends/com_backend.py) with placeholders for meta-tools. Update CLAUDE.md to document the premium drawing rules and standard workflow. Misc: add examples, references, and tests for premium features, and minor server/backend adjustments.
Introduces an engineering/ package that lets MCP clients produce production-
quality technical drawings without LLM-improvised geometry. Adds 7 new MCP
tools, an 8-step pre-completion validator, and an automatic linetype + layer
bootstrap on drawing_new so CENTER/HIDDEN/PHANTOM and standard mechanical
layers (GEOMETRY, DIM, HATCH, TEXT, TITLEBLOCK, ...) are always available.
- engineering/gear.py: pure-Python involute math, 40+ points/flank, helical
+ spur front views, deterministic section A-A (no helix lines in section).
- engineering/keyway.py: DIN 6885 keyway table + parametric keyed bore /
side-section drawers.
- engineering/titleblock.py: ISO 7200 A3 (420x297 mm) title block; title
text is written verbatim to the title cell.
- engineering/validator.py: DrawingValidator with 8 checks (real DIMENSION
entities, no hidden orphans, linetypes loaded, bore + keyway in both
views when expected, title consistency e.g. forbid 'HELICAL SPUR GEAR'
when helix_angle truthy, saved-to-disk + file-on-disk, screenshot
exported, dwg path returned).
- backends/ezdxf_backend.py: AutoCAD-only linetype fallback (HIDDEN,
BORDER families) so ezdxf doc gets correct patterns when linetype_load
is called for names ezdxf.tools.standards does not ship.
- backends/base.py: concrete helpers set_layer_active / ensure_linetypes.
- server.py: Section 12 with 7 engineering tools, drawing_new(bootstrap=
True), and an ENGINEERING WORKFLOW block in the MCP instructions string.
- 30 new tests covering gear math, DIN6885 lookup, ISO A3 layout, title
verbatim, and every validator check. Tool count 89 -> 96.
