If you're building or evaluating CAD models in ForgeCAD and need more than eyeballing the output, this skill runs a structured grading workflow. It converts your design brief into a checklist, renders the model from multiple angles, runs targeted inspections when fit or internals matter, and assigns a 0-10 score with evidence. The rubric covers requirement fit, geometry, mechanical plausibility, and model health, plus hard caps if the code doesn't run or critical features are missing. Useful for QA, benchmarking prompt-generated models, or comparing design iterations when you want a defensible number instead of vibes.
npx -y skills add kostard/forgecad-public-kit --skill forgecad-model-grader --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
wshobson/agents
github/awesome-copilot