mcp · Global

CH-M-I-M-AGENT

Bo cham diem AI Agent dung Skill/Tool cho AutoCAD - Civil 3D - Revit BIM cau. 48 case, 6 hard gate, 197 test cho chinh bo cham.

Claim this listing

Represent CH-M-I-M-AGENT? Send a claim for review so we can discuss listing updates. Claims do not automatically grant ownership or verification.

Data offered llm-inferred: SQLite agent-task scoring (16 cases, against a live in-memory database) · Autodesk Revit 2026 bridge-BIM agent scoring (36 cases) · AutoCAD and Civil 3D agent scoring (12 cases, simulated) · Post-run snapshot verification of agent end state
Formats llm-inferred: JSON
Access docs-verified: Free, MIT licence. Basic SQLite and simulated runs need no install, API key or external software; the Claude adapter needs ANTHROPIC_API_KEY, and live Revit mode needs Revit 2026 with revit-mcp.

Use cases llm-inferred: Check whether a Revit or Civil 3D automation agent actually changes the model before trusting its SUCCESS reports on BIM deliverables · Gate-test agent skills for destructive actions such as bulk deletes before they touch project files · Compare candidate LLMs on CAD/BIM task reliability using pass^k across repeated runs
Audience llm-inferred: BIM and civils engineers evaluating AI automation · AI tooling developers building CAD/BIM agent skills · Digital engineering or IT leads choosing agent tooling

Constraints llm-inferred: Benchmark harness and grader, not a dataset or hosted API · AutoCAD and Civil 3D pack is simulated only; no live mode is documented · Revit live mode needs Revit 2026, revit-mcp and a disposable model copy; writes are double-gated · Public suite is small (64 cases) and the public SQLite pack did not separate models · Public Revit pack is 36 cases; a private companion Revit pack is referenced but not public · ENTRY metadata (48 cases, 197 tests) does not match DOCS (64 cases, 351 checks) · No UK-specific data, standards or regional coverage documented · Arbitrary C# via send_code_to_revit is not a security sandbox

Licence: MIT
Authentication: API key via env: ANTHROPIC_API_KEY (auto-extracted from README)
GitHub stars: 3
Last commit: 24/08/2026
Last successful page check: 14/10/2026
Page health: healthy
Package: pypi:anthropic
Verification: Not independently verified

Auto-extracted, unverified fields: tools, install, auth. Check upstream documentation.

Daily discovery 2026-10-15: GitHub search; verified via GitHub API. Tools/install/auth auto-extracted from README.

Metadata observed 14/10/2026 · daily-2026-10-15 (github-search)