Anthropic cut roughly 80% of Claude Code’s system prompt
Anthropic says it trimmed about 80% of the Claude Code system prompt for its newest models, Opus 5 and Fable 5. The company’s post on X links to a breakdown of what the change means for writing system prompts, skills, and Claude.MD files.
We covered some of those takeaways in our piece on the updated context engineering rules for the Claude 5 generation. The post went up July 2; Anthropic introduced Opus 5 on July 24.
Up next
Anthropic rewrites its context engineering rules for Claude 5
Anthropic’s new Claude 5 context engineering guidance drops rigid rules for the model’s own judgment: an early Claude Code prompt banned multi-line comments, now it just asks the model to match surrounding code style. Tool examples are out; interface design is in.
Related stories
- 44 KB of Claude Code’s classifier prompt never gets printed
- A hardcoded instruction in Claude Code 2.1.219 and 2.1.220 blocks Opus 5 from spawning sub-agents
- Pinned Claude Code subagents fall through to Fable
- Claude Code cut over 80% of its system prompt, scores held
- Delete “think carefully” from your Opus 5.5 prompts
- “Verify twice” wastes tokens on Opus 5, Anthropic says
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.