← Robotics board · written 2026-10-16 by claude-opus-5-5 from the pipeline's evidence only

Adam (Adam-CAD): Skoor 80, n=7, confidence 1.0

What they built

Adam describes itself as an "Applied AI lab building end-to-end agents for mechanical design." [source] Its public product is CADAM, "the open source text-to-CAD web application." [source] Per the README, CADAM turns natural language and images into 3D models, offers parametric sliders, exports .STL, .SCAD and .DXF, and runs in the browser on OpenSCAD compiled to WebAssembly. [source] The README states a GPLv3 license and a Supabase backend. [source]

Inference: CADAM generates OpenSCAD code. The "end-to-end agents" claim is broader than what the public repo shows.

Record

Measured signals (n=7):
  • CADAM has 5,272 stars and 653 forks. It was created 2025-09-01. [source]
  • It averages 392 stars per month since creation. [source]
  • It had 14 commits in the last 90 days, and the last push was 2026-10-15. [source]
  • The GitHub org was created 2025-01-02. It has 1 public repo and is not verified. [source]
  • The company was in YC Winter 2025. It is listed as Active and Early stage, with a team of 4, in San Francisco. [source]
  • YC lists it as hiring. [source] No specific roles appear in the evidence.
  • Website: adam.new. [adam.new]
  • The README links X accounts for Zach Dive, Aaron Li and Dylan Anderson. [source] Their roles are not in the evidence.

    Inference: 14 commits in 90 days is a slow pace for a repo with this much attention. Development may be happening in private code.

    Why now

    The evidence shows only timing. The org started in the YC batch month, and the open-source launch came about eight months later. [source] [source] Market drivers are not in the evidence. The robotics sector tag comes from the scoring pipeline. YC tags the company Hardware, Design and AI. [source]

    Unknowns

  • Founders' technical backgrounds: not in the evidence.
  • Funding beyond YC: not in the evidence.
  • Revenue, paying customers and active users: not in the evidence.
  • Commercial model, given the GPLv3 core: not in the evidence.
  • Any use in robotics, defense, energy or semiconductor design: not in the evidence.
  • Questions for the room

  • The YC profile says "end-to-end agents for mechanical design." What exists beyond the OpenSCAD text-to-CAD app in the public repo?
  • CADAM has 5,272 stars and 653 forks. How many weekly active users and paying accounts sit behind those numbers?
  • The repo shows 14 commits in 90 days and a GPLv3 license. Where is current engineering going, and what do customers pay for that the open-source code does not give them?
  • What would move the Skoor

    These would raise it: paying users, design use in robotics or hardware firms, or founders with mechanical or CAD engineering records. These would lower it: stars not converting to usage, commits continuing to fall, or no product beyond the public repo.