Iterating a Landing Page Across Claude, Codex and Kimi Until It Ships
aryanXmahajan · x · 2026-09-05
The author's LLM philosophy: don't give up after one bad result — treat the model like an old TV, smack it, change the angle, switch models, and keep going. Inspired by a walk through an Apple Store, he built an Apple-style lead magnet page for his LinkedIn Ghostwriter's 11 Claude skills, each card opening one skill. No 39-page agent architecture: he just passed the page back and forth between Claude, Codex and Kimi, telling each what still looked bad until it looked right.
More from coding & agent
- Astra bypasses harness interface, figures out how to spawn Codex sub-agents directly — arthurcolle · 2026-09-05
- Open-source pipeline classifies phone screenshots, then analyzes and searches the web — TheMoonMidas · 2026-09-05
- YouTube Tutorial Shares a Free MiniMax-H3 Workflow — solomars3 · 2026-09-05
- SemiAnalysis and vLLM launch AgentX, a benchmark for real multi-turn agentic traffic — AccBalanced · 2026-09-05
- Zilliz CTO on Notion's Vector Infra: 90% Cost Cuts and the Platform Hiding Under One AI Feature — J_Luan_ · 2026-09-05
- A Commerce Agent Needs Spending Boundaries at the Card Layer, Not the Prompt — NoFunnyMan · 2026-09-05