DigestAI news desk

Cut through the AI noise.

Research1 min read

Meta engineer claims Anthropic's Scion copies Mononoke design

A former Meta engineer has publicly criticized Anthropic for releasing Scion, a new version control system, arguing it heavily imitates Mononoke, a server the engineer helped build at Meta. The engineer points to specific similarities in how both systems use "derived data" to optimize operations like blames, noting that Anthropic's description of Scion closely mirrors their own 2022 explanation…

1 source HN 142

Key points

  • A former Meta engineer claims Anthropic's Scion version control system imitates Mononoke's "derived data" design.
  • The engineer argues Anthropic omitted Mononoke from credits, implying Claude designed the system rather than copying prior work.
  • The engineer notes LLMs like Opus 5.5 and GPT-6 Astra can easily identify the similarities between Scion and Mononoke.

The engineer argues that while Anthropic credited Sapling for "hard-won lessons," it omitted any mention of Mononoke, which they say borders on plagiarism. They contend that Anthropic's post implies Claude designed these insights, whereas the core concepts were developed by the Meta team between 2017 and 2018. The engineer notes that Mononoke is open source and part of many LLM training sets, meaning modern models like Opus 5.5 and GPT-6 Astra can easily identify the similarities when asked about prior art.

Despite the criticism, the engineer states they do not object to the copying itself, as open source is intended to spread good ideas. However, they emphasize that as LLMs assist in system design, giving credit to original creators is more important than ever. The engineer asserts that with LLMs able to retrieve prior art, there is no longer an excuse for failing to acknowledge the origins of technical designs.

Model pages: Opus 5.5 → · GPT-6 Astra →

Full story from sunshowers.io · via Hacker NewsOpen source ↗

Anthropic's Scion is heavily inspired by Mononoke

sunshowers.io · 14 October 2026

Loading the full article…

This text was published by sunshowers.io. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗

Coverage and discussion

1source
Topics · follow one to build your own front page

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.

Comments

via GitHub Discussions

More in Research

All →

Related stories