Claude Fable 5.2 coding edits follow prompt ownership wording
Original title: Quoting Prompting Claude Fable 5.2
Calling the codebase the model’s led it to merge repeated code in about half of small-task tries, versus under a tenth otherwise.
Simon Willison quotes Anthropic’s notes on prompting Claude Fable 5.2. When a prompt said nothing about ownership or treated the code as someone else’s, the model tended to extend the existing structure, including by repeating code. Telling the model that the codebase was its own made structural cleanup much more common in Anthropic’s small coding-task tests.
Unconfirmed: The feed snippet is truncated, so only the quoted Anthropic testing result is available.
- The cited tests covered small coding tasks, not large production repositories.
- Without an ownership instruction, the model more often added to existing structure.
- The suggested sentence tells the model the codebase and its design are its own.
- Vendors
- Anthropic
- AI models
- Claude Fable 5.2
From the original article
Claude Fable 5. 2 responds closely to what the prompt says about who owns the code it works on. In Anthropic’s testing on small coding tasks, when a prompt said nothing about ownership or presented the code base as someone else’s, the model tended to add to the existing structure, even when that repeated existing code.
simonwillison.net
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.