Expired90% confidenceEventTime unknown
Theo ran six Agents simultaneously for testing: two Gemini, two Claude Code (Opus 4.5), and two GPT 5.2 (Codex), testing the effect with and without a frontend design skills file respectively.
1
Sources
90%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026(expired)
Sources
AI前端设计实测：Opus 4.5靠一个Markdown技能文件逆袭登顶
bilibili一摩尔炸鸡翅
Related Entities
Related Claims
UnverifiedIn a domino physics simulation test, Opus 4.5 completed five formation designs with a single prompt, while GPT-5.2 Codex required two prompts to reach a basically usable state.74% similarUnverifiedThe benchmark tested leading AI models including Claude, GPT series, and Gemini.73% similarUnverifiedIn actual tests of UI component generation tasks, Composer's speed far surpassed competing products, with Claude ranking second and GPT-5 clearly lagging behind.73% similarUnverifiedEven though Theo explicitly requested that skill files not be used, Codex's system priority still forced the frontend design skill to be invoked, meaning GPT 5.2's bare run results may have already been enhanced by the skill file.73% similarUnverifiedTheo estimated that the GPT-5 series has approximately 10 built-in design templates, Opus has around 20, and Gemini has only 1 to 2.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/20546API
curl https://kongchang.com/api/v1/knowledge/claims/20546MCP
get_claim(id=20546)