OpenAI 发布 GPT-6 Astra 提示词指南，含 slop 词屏蔽清单

OpenAI 的模型文档详细说明了 GPT-6 Astra 在哪些情况下容易出现非预期行为，以及开发者如何规避这些问题。

据 OpenAI 介绍，GPT-6 Astra 比 GPT-5.6 Sol 更频繁地提出澄清性问题，而不是自行做出假设，因此它是一位“更高效的协作者”。但代价是，该模型有时会在用户期望它继续推进的地方停下来。

为了促使模型更具主动性，OpenAI 建议在提示词中告诉模型从上下文推断用户的“意图”，并表现出“行动偏好”。像“你能……”“我想……”或“帮我……”这类表述应被视为行动指令，而非引发追问的邀请。

You should infer the user’s intent and the scope of the task from the instructions and the conversation context so far. Your task is to demonstrate a tendency to act and to follow through on the user’s intended task until completion. If the user expresses the intention to complete new work or resolve an existing problem, continue working persistently until the user’s intended goal is achieved. Work independently toward the user’s goal (e.g., create isolated work trees/checkouts, resolve merge conflicts, perform read-only actions, create draft PRs, etc.), unless the actions are clearly destructive or irreversible.

模型应等到已经准备好一个具体、可审查的结果后，再请求用户批准。OpenAI 以这种方式引导模型：“用户应该批准的是一个具体、可审查的结果。”基于假设性风险的主动警告、免责声明或安全检查清单应从提示词中删除。

GPT-6 Astra 比其前代模型能更好地遵循较长的指令，但对上下文也更为敏感。技能文件（如 AGENTS.md）中不清晰或相互矛盾的指令可能导致模型停止工作或意外偏离方向。OpenAI 建议审计模型可访问的所有技能文件和上下文文档，并明确赋予用户指令最高优先级。

OpenAI 还推荐了一个调试提示词，迫使模型明确指出是哪一个技能文件、并引用具体哪一条指令导致它暂停或改变方向。这有助于开发者将意外行为追溯到其根源。

If a skill causes you to ask for permission or confirmation, pause, leave requested work unfinished, or diverge from the user's intent, name and link to the exact SKILL.md file you read, quote the relevant instruction, and briefly explain how it applies. Distinguish explicit skill requirements from your interpretation of guidelines.

“Delve”、“foster”、“leverage”以及其他那些“口水词”

GPT-6 Astra 倾向于用列表、表格和 Markdown 格式来组织回答，并且会在不同会话中重复使用相同的措辞。OpenAI 对塑造模型的写作风格有具体的指导。如果你想要散文式的文字，就要明确告诉模型用平实的语言和主动语态写出简洁的段落。

By default, use clear, concise paragraphs, each developing a single main idea. Use lists only if the information is truly parallel, sequential, or more easily comparable, and avoid nested lists unless the hierarchy cannot be clearly expressed in prose. Use simple, straightforward language: familiar words, concrete examples, and precise verbs. Favor the active voice and direct statements. Make the main point clear early on, then expand on it with the explanation and details the reader needs. Let each sentence build on the previous one. Develop the points that are important and provide enough evidence to be useful.

一份典型的 AI 措辞黑名单也能派上用场。OpenAI 称这些为“口水词”。像“exact-head checks”或“editorial-row layouts”这类生造的连字符复合词也应避免使用。模型应该说明自己在做什么，而不是罗列自己不会做什么。

Avoid using slop words or phrases such as “Conclusion:” in conclusions, “delve into,” “promote,” “use/leverage,” “it’s worth noting,” “what’s important is,” “Question? Answer,” or “This isn’t about X. It’s about Y,” “really/truly,” or compound descriptions and hyphenated adjectives. Do not use summary closing statements such as “In short:...,” “The simplest mental model is:...”. State the intended action directly. Avoid mentioning what you won’t do, what remains unchanged, or how you’ll separate or categorize results. Do not use contrastive phrasing such as “X, not Y” or “X—not Y,” which introduces an unsolicited alternative that the user did not ask for. Avoid made-up compound terms like “exact-head checks” and “editorial-row layouts,” vague qualifiers, and stock transitions; use simple verbs and prepositions to directly express the actual relationship.

对于技术写作，OpenAI 建议只在真正有帮助的情况下才使用术语：“优先使用平实的语言而非术语，只有在有助于向用户阐明某个想法或你的工作时，才引用技术细节。”

子智能体委派不足，测试规模膨胀

该模型可以将工作移交给并行运行的子智能体，但实际这样做的频率低于预期。OpenAI 表示，开发者应当明确说明何时以及多大程度地让它进行委派。智能体之间的消息也可能包含语法或空格错误。

在编码任务中，GPT-6 Astra 会在收尾前运行全面的测试。对于小改动而言，这可能意味着测试套件的规模与实际工作量严重不成比例。OpenAI 建议开发者告诉模型，仅在出现新的失败或未解决问题时才重新运行测试。

这些提示词的更详细版本可在GPT-6 Astra 模型文档页面上查看。想要切换到 GPT-6 Astra 的开发者可以使用 Codex 配合OpenAI Docs 技能来自动应用推荐的更改：$openai-docs migrate this project to GPT-6 Astra