๊ฐ€์ž… ํ›„ ์ดˆ๋Œ€ ๋งํฌ๋ฅผ ๊ณต์œ ํ•˜๋ฉด ๋™์˜์ƒ ์žฌ์ƒ ๋ฐ ์ดˆ๋Œ€ ๋ณด์ƒ์„ ๋ฐ›์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

Artzy
@heyartzy
AI & Tech Creator | sharing my takes on AI, tools | KOL, community & outreach @nabulines ๐Ÿ‘๏ธโƒค| CCP @imagineart_x
๊ฐ€์ž… February 2023
1.7K ํŒ”๋กœ์ž‰ ์ค‘    2K ํŒฌ
THIS IS HUGEโ— : @antigravity now puts Claude Opus 5.5, Sonnet 5.5, and GPT-OSS-120B in the same workspace โ€ผ๏ธ With a non-trial Google AI Pro account, you can access all three alongside Gemini. The interesting part is what you can hand from one model to another: ๐Ÿ‘‡ Let Gemini map the repo, find the relevant files, and handle routine edits. Then bring Claude into a specific decision: why a test keeps failing, whether an auth change opens a hole, or which of two fixes fits the codebase. repo exploration โ†’ focused brief โ†’ Claude review โ†’ tests Thatโ€™s a workflow worth testing when Claude quota is limited. Give it a smaller, clearer problem before spending the allowance on rediscovering the project. The handoff is where this can break. Leave out an important dependency or constraint, and Claude can miss the problem too. Track quota used, corrections needed, and whether the change passes. That tells you where each model earns its place in the workflow.
๋” ๋ณด๊ธฐ