Anthropic cuts Claude Code prompts by 80%, but Opus 5 adds 72% back
量子位 · wechat · 2026-07-27
Anthropic says it cut more than 80% of Claude Code’s old system prompt rules as it shifted from an oversized “employee handbook” to lighter context engineering. But a community trace of the actual prompts sent to different models found a more nuanced story: Opus 4.7 had 15,225 characters, Opus 4.8 dropped to 4,467, and Opus 5 climbed back to 7,694—about 72% longer than 4.8, though still roughly half of 4.7.
The article explains why both numbers can be true:
- The big reduction happened in the transition to the new prompt system for newer models.
- Opus 5 got extra instructions back because it behaves more proactively, uses sub-agents more often, and tends to produce longer outputs.
- The new additions mainly cover delivery/workflow control and correction behavior: scope management, progress reporting, when to stop, and how much to self-correct.
For developers using Claude Code or building agents on Claude, Anthropic’s guidance is to keep CLAUDE.md lean, split long instructions into Skills, remove duplicate rules, reduce tool-call examples, and provide executable references like tests or existing code instead of verbose prose.
Related event: Anthropic Slashes Claude Code System Prompts Without Losing Performance(2 posts)→
More from coding & agent
- Port22 asks whether phone-first coding agents should start runs or only approve them — casualhermit · 2026-07-27
- A new agent memory framework says context management is an architecture problem — Gaurav Dadhich · 2026-07-27
- Agent performance tanked after one team loaded 600+ skills at startup — dbreunig · 2026-07-27
- Chinese AI founder gives a 40-minute class on agent swarms at a $20B company — aftahi_ai · 2026-07-27
- Codex and Claude Code become the punchline in an AI-math conjecture joke — charles_irl · 2026-07-27
- A trader says Claude-vibe-coded market-making bot loses 0.75 bps per trade — alejandroll10 · 2026-07-27