release: v1.5.0 — discovery layer, honesty debt closed, Wave A

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>
This commit is contained in:
Umut CAN
2026-08-06 09:45:20 +03:00
co-authored by Claude Opus 5
parent d5a0ff5109
commit b03a7615b1
107 changed files with 32640 additions and 4810 deletions
+60 -4
View File
@@ -14,8 +14,9 @@ import time
from datetime import UTC, datetime
from pathlib import Path
from backends.capability import UnsupportedCapabilityError
from benchmarks.adapters.base import BenchmarkAdapter, TaskResult
from benchmarks.tasks_v2 import TASKS_V2, TaskSpec, task_by_id
from benchmarks.tasks_v3 import DEFAULT_MATRIX, MATRICES, TaskSpec, task_by_id
def _load_registry() -> dict[str, dict]:
@@ -78,6 +79,7 @@ async def run_tasks(
artifact_dir: str | Path,
*,
timeout: float = 30.0,
matrix: str = DEFAULT_MATRIX,
) -> dict:
destination = Path(artifact_dir).resolve()
destination.mkdir(parents=True, exist_ok=True)
@@ -90,6 +92,12 @@ async def run_tasks(
result = await asyncio.wait_for(adapter.run_task(task), timeout=timeout)
except TimeoutError:
result = TaskResult(task.task_id, "timeout", 0.0, f"Exceeded {timeout}s")
except UnsupportedCapabilityError as exc:
# "Your server got this wrong" and "your engine cannot reach
# this" are different findings about a competitor. The enum has
# carried `unsupported` since v2 for exactly this, but a
# refusal is still an exception and used to land as `fail`.
result = TaskResult(task.task_id, "unsupported", 0.0, f"[{exc.capability}] {exc}")
except Exception as exc:
result = TaskResult(task.task_id, "fail", 0.0, str(exc), stderr_summary=str(exc))
result.duration_ms = round((time.perf_counter() - started) * 1000, 2)
@@ -108,7 +116,11 @@ async def run_tasks(
weight_by_id[item.task_id] * item.score for item in results if item.status != "unsupported"
)
return {
# The *report* schema is unchanged — same keys, same enum — so the
# chart renderers still read v1.4's published files. `matrix` says
# which task set ran, which is the thing that actually moved.
"schema_version": "2.0",
"matrix": matrix,
"adapter": adapter.name,
"adapter_metadata": adapter.metadata(),
"git_sha": _git_sha(),
@@ -137,6 +149,12 @@ async def run_tasks(
def _parser(registry: dict[str, dict]) -> argparse.ArgumentParser:
parser = argparse.ArgumentParser(description=__doc__)
parser.add_argument("--list", action="store_true", help="List the fixed benchmark tasks")
parser.add_argument(
"--matrix",
default=DEFAULT_MATRIX,
choices=sorted(MATRICES),
help="Task set to run. v2 stays addressable so a v1.4 report can be reproduced.",
)
parser.add_argument(
"--server",
default="autocad-mcp-pro",
@@ -148,14 +166,35 @@ def _parser(registry: dict[str, dict]) -> argparse.ArgumentParser:
parser.add_argument("--artifact-dir", default="benchmarks/results/latest")
parser.add_argument("--timeout", type=float, default=30.0)
parser.add_argument("--json", action="store_true", help="Print machine-readable JSON")
parser.add_argument(
"--publish",
type=Path,
help=(
"Also write the report here with artifact paths reduced to filenames. "
"How results/published/*.json are produced — it was a manual edit before."
),
)
return parser
def _sanitize_for_publication(report: dict) -> dict:
"""Drop absolute paths from artifacts, keep the hash that identifies them.
A published report should not carry the run machine's directory layout;
the sha256 is what makes the artifact checkable, and that stays.
"""
published = json.loads(json.dumps(report))
for result in published["results"]:
for artifact in result["artifacts"]:
artifact.pop("path", None)
return published
def main() -> None:
registry = _load_registry()
args = _parser(registry).parse_args()
if args.list:
for task in TASKS_V2:
for task in MATRICES[args.matrix]:
print(f"{task.task_id}\t{task.category}\t{task.description}")
return
@@ -165,10 +204,27 @@ def main() -> None:
raise SystemExit(
f"{args.server} supports backends {supported_backends}, not {args.backend!r}"
)
tasks = [task_by_id(task_id) for task_id in args.task] if args.task else list(TASKS_V2)
tasks = (
[task_by_id(task_id, args.matrix) for task_id in args.task]
if args.task
else list(MATRICES[args.matrix])
)
adapter = _make_adapter(entry, args.backend)
report = asyncio.run(run_tasks(adapter, tasks, args.artifact_dir, timeout=args.timeout))
report = asyncio.run(
run_tasks(adapter, tasks, args.artifact_dir, timeout=args.timeout, matrix=args.matrix)
)
rendered = json.dumps(report, ensure_ascii=False, indent=2, sort_keys=True)
if args.publish:
args.publish.parent.mkdir(parents=True, exist_ok=True)
args.publish.write_text(
json.dumps(
_sanitize_for_publication(report), ensure_ascii=False, indent=2, sort_keys=True
)
+ "\n",
encoding="utf-8",
newline="\n",
)
print(f"Published: {args.publish}", file=sys.stderr)
print(rendered)