by capskov
AutoCAD callout annotation Skill for AI agents: ordinary LEADER/MTEXT styling, bounded pairwise layout, and bilingual BZYD AutoLISP manual adjustment.
# Add to your Claude Code skills
git clone https://github.com/capskov/cad-callout-pair-layoutGuides for using ai agents skills like cad-callout-pair-layout.
See how cad-callout-pair-layout compares with popular alternatives.
cad-callout-pair-layout is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by capskov. AutoCAD callout annotation Skill for AI agents: ordinary LEADER/MTEXT styling, bounded pairwise layout, and bilingual BZYD AutoLISP manual adjustment. It has 0 GitHub stars.
cad-callout-pair-layout's catalog security scan is still queued. You can run an instant dependency and prompt-injection check now with the "Scan for vulnerabilities" button above.
Clone the repository with "git clone https://github.com/capskov/cad-callout-pair-layout" and add it to your Claude Code skills directory (see the Installation section above). cad-callout-pair-layout ships a SKILL.md manifest, so compatible agents can discover and load it automatically.
cad-callout-pair-layout is primarily written in Python. It is open-source under capskov on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh cad-callout-pair-layout against similar tools.
No comments yet. Be the first to share your thoughts!
Available after the catalog security scan passes.
⚠️ Third-Party Software Notice
This skill is third-party open-source software developed and hosted independently on GitHub. SkillsLLM is an informational directory and does not control or maintain the underlying repository.
Any security checks, ratings, or warnings displayed by SkillsLLM are automated and limited in scope. They do not constitute a security certification or guarantee that the software is safe, error-free, or free from malicious code, vulnerabilities, compromised dependencies, or prompt-injection risks.
Review the source code, permissions, dependencies, and configuration before installing or running any third-party skill. Use is at your own risk. To the maximum extent permitted by applicable law, SkillsLLM is not liable for losses arising from third-party software.
The deep catalog scan for this skill is still queued. Run an instant dependency check now instead.
See comparison
Source preview v0.6 uses Agent-led planning. 完整中文规则, planning guide, input/acceptance, transport contract. The unchanged v0.3 pairwise solver is an optional local repair tool; its fixed angle/extension limits do not constrain Agent planning. An unresolved ID returns to the Agent for judgment, not necessarily to the user.
Use ordinary three-vertex LEADER associated with MTEXT and an actually yellow anchor dot, not MLeader. Preserve approved content, quantities and base geometry. B–C follows frame horizontal, with approved text gap, a strictly obtuse B angle and no leader crossing its own text. Read actual task-approved style overrides; historical metre-based dimensions are examples, not universal defaults. Automatic scale selection/resizing remains unimplemented.
Verify native AnnotationOffset, Dimtad and regeneration, not just stored vertices. See stable association/tail. Keep native objects editable and new layers unlocked; do not explode or rasterize them. On an initial field trial, edit one MTEXT in the native editor and restore it, then read back anchor/association/horizontal nonzero tail. Significant text-width changes need coordinated tail maintenance; arbitrary edits are not promised automatic equal-width tails.
Discover actual runtime schemas and uniquely bind process/full path/backend before writes. Unknown writes are queried by their original request, never replayed under new IDs. Report native capability gaps; task-authorized alternatives do not bypass refusals or fences. This Skill ships no CAD writer, arbitrary-code adapter, fonts or engineering DWG. Foundation supplies reusable operations; the Agent/Skill supplies engineering presentation judgment.
The accepted 35-callout field example displays the second-area PDF crop and its pm.ctb PDF. Annotation writes/readback/save used Foundation MCP; manual editing used CAD UI; explicitly authorized COM performed plotting. It is a local-case validation, not all-drawing or all-version certification. See evidence. The optional BZYD manual utility remains separate and is not an Agent execution backend.
中文 · Install · Compatibility · Evidence
Agent-led CAD callout planning, with small geometry tools for reliable repetitive checks. Source preview v0.6 reads the inner frame first, plans regions/order/openings, then applies and reviews native results. The v0.3 pairwise solver remains available for local repair.
| Task | Components | Boundary |
|---|---|---|
| Supported basic CAD only | cad-agent-foundation | This Skill is optional; core is currently a source preview. |
| This project's callout style/placement | Foundation + this Skill | Full style/placement integration is not certified on Foundation v1 alone; inspect native gaps. |
| Distress/demolition | Foundation + this Skill + cad-distress-annotation | Distress rules provide professional scope; public rules preview v0.1.0-rc.1 is available separately. |
| Offline proposed geometry only | This solver + Python/Shapely | No CAD/Foundation needed; calculation does not prove native placement. |
Defaults/native history come from metre-based road drawings. Other professions, planting/species labels and millimetre styles require adaptation and are not validated. The AI Skill/solver has no native writer or automatic installer; optional manual Lisp is separate. No fonts, engineering DWGs or legacy COM helpers are bundled; the selected second-area PDF and crop are display evidence. Try the synthetic example.
The new 35-callout field trial separates native MCP annotation evidence from authorized COM plotting. Historical evidence and remaining limits are listed in validation.
CAD Agent Foundation supplies supported reusable native operations and execution governance. This Skill supplies style parameters/placement rules; missing operations do not appear by installing it. The related cad-distress-annotation supplies engineering scope/classification. Repository: cad-callout-pair-layout.
The anonymous display case is a comprehensive historical project Python/AutoCAD COM workflow, not Foundation-only execution proof. Reproduction requires the stated components, verified extra native operations and cleared inputs; no automatic CAD replay dataset is included.
Owned source/instructions: Apache-2.0; dependency terms remain separate. Notices. Relevant terms: AutoCAD, CAD annotation, ordinary LEADER, MTEXT, callout layout, CAD Agent Skill, MCP / Model Context Protocol. This Skill is not an MCP server or Autodesk product.
Optional manual utility: BZYD, English/Chinese ANSI editions. It adjusts existing callouts without Foundation/Python. AutoCAD2026 Chinese historical tests exist; AutoCAD2022 and the translated English native path remain unverified.
Structure and CAD2022/2026 support by component. Native dimensions are a future independent Skill, not a reason to widen this package's current support claims.
