FULL STORY
GPT-6 Astra Prompting: From Critique to Official Guide
After a developer argued legacy skills and bloated prompts now hurt GPT-6 Astra, OpenAI shipped an official prompting guide confirming rules should be drastically trimmed.
2026-09-05 ~ 2026-09-06 · 2 episodes · 7 posts
Episode 1 · OpenAI Devs Say Legacy Skills Now Hurt Performance, Urge Rebuilding for GPT-6 Astra (2026-09-05, 4 posts)
OpenAI developer pvncher argues in 'Rethinking skills and prompts for GPT-6 Astra' that scaffolding built for weaker models has become bloated and even counterproductive, and that coding-agent best practices are rapidly shifting with GPT-6 Astra.
- Rethinking skills and prompts for GPT-6 Astra: coding agent best practices are changing fast — pvncher · 2026-09-05
- Rethinking skills and prompts for GPT-6 Astra: old agent scaffolding is fading — EyalToledano · 2026-09-05
- Rethinking skills and prompts for GPT-6 Astra: coding agent best practices are shifting — iamrobotbear · 2026-09-05
- OpenAI researcher: old Skills now hurt GPT-6 Astra — audit and clean up your AGENTS.md — udmrzn · 2026-09-06
Episode 2 · OpenAI's GPT-6 Astra prompting guide: trim your rules (2026-09-05, 3 posts)
OpenAI released an official prompting guide for GPT-6 Astra, advising users to drastically trim SKILL and AGENTS.md rules and including a built-in prompt to curb the model's tendency to run excessive tests.
- GPT-6 Astra guide ships an official prompt to curb the model's excessive test-running habit — JeremyNguyenPhD · 2026-09-05
- OpenAI's GPT-6 Astra prompt guide: strong model means you should trim your SKILL and AGENTS.md rules — xiaohu · 2026-09-06
- OpenAI's GPT-6 Astra prompting guide: trim your SKILL and AGENTS.md rules, the model is that strong — xiaohu · 2026-09-06