Microsoft trains a Markdown skill and lifts GPT-5.5 scores
SkillOpt treats an agent instruction file like trainable parameters, improving benchmark scores without changing model weights.
2 verified stories covering Prompt engineering, product updates and industry developments.
SkillOpt treats an agent instruction file like trainable parameters, improving benchmark scores without changing model weights.
OpenAI has updated its official GPT-5.5 prompt guide, urging developers to stop reusing prompts from GPT-5.2 or 5.4, as they hinder the new model's performance. The guide recommends a fresh start with a new seven-part structure and notes a default reasoning effort change from high to medium.