Claude system prompts as a git timeline
Simon Willison converted Anthropic's published Claude system prompts into a git repository with timestamped commits, enabling visual tracking of how the prompts have evolved across model versions and families.
Modelwire context
Analyst takeWillison's project isn't just a curiosity for prompt engineers — it creates a public audit trail for how Anthropic's behavioral guardrails and model instructions shift over time, without Anthropic having to announce those changes explicitly. That's a form of external accountability that Anthropic didn't build and may not have anticipated.
Anthropic has had a busy week of product and policy moves: Claude Design launched on April 17th, the cybersecurity-focused Claude Mythos Preview dropped the same day as a likely government relations play, and Schematik's hardware AI tool drew Anthropic investment interest on April 18th. What's notable is that none of those releases came with detailed disclosures about how the underlying system prompts or behavioral constraints were adjusted. Willison's git timeline makes those gaps visible in retrospect, turning Anthropic's own published materials into a changelog the company never formally offered.
Watch whether Anthropic responds by formalizing its own versioned prompt changelog, or whether it quietly reduces the granularity of what it publishes — either response would signal how the company weighs external scrutiny against controlled messaging.
This analysis is generated by Modelwire’s editorial layer from our archive and the summary above. It is not a substitute for the original reporting. How we write it.
MentionsAnthropic · Claude · Simon Willison · GitHub
Modelwire summarizes — we don’t republish. The full article lives on simonwillison.net. If you’re a publisher and want a different summarization policy for your work, see our takedown page.