{"@context":"https://schema.org","@type":"NewsArticle","generatedAt":"2026-09-06T14:20:08.399Z","headline":"OpenAI 发布 GPT-6 Astra 提示词指南，含 slop 词屏蔽清单","description":"OpenAI 在模型文档中说明 GPT-6 Astra 相比 GPT-5.6 Sol 更常提出澄清问题、对上下文更敏感，并给出让模型更主动、审计 AGENTS.md 等技能文件、控制写作风格、约束子智能体委派和测试规模的提示词建议。","url":"https://www.aioga.com/news/cmtofayvo02uoroht238ucfjw/","mainEntityOfPage":"https://www.aioga.com/news/cmtofayvo02uoroht238ucfjw/","datePublished":"2026-09-05T13:31:04.000Z","dateModified":"2026-09-05T13:31:04.000Z","inLanguage":"zh-CN","publisher":{"@type":"NewsMediaOrganization","name":"Aioga","url":"https://www.aioga.com"},"citation":["https://the-decoder.com/openai-shares-prompting-tips-for-gpt-6-astra-including-a-blocklist-of-slop-words","https://aihot.virxact.com/items/cmtofayvo02uoroht238ucfjw"],"canonicalUrl":"https://www.aioga.com/news/cmtofayvo02uoroht238ucfjw/","directAnswer":{"@type":"Answer","text":"OpenAI 模型文档介绍了 GPT-6 Astra 的行为特点与提示词建议。文档称，该模型比 GPT-5.6 Sol 更常提出澄清问题，也更重视上下文；同时提供了推动模型主动执行、审计技能文件、控制写作风格、约束子智能体委派和测试规模的方法。","url":"https://www.aioga.com/news/cmtofayvo02uoroht238ucfjw/","dateCreated":"2026-09-05T13:31:04.000Z","author":{"@type":"Organization","@id":"https://www.aioga.com/authors/aioga-editorial/#editorial-team","name":"Aioga Editorial Team","url":"https://www.aioga.com/authors/aioga-editorial/"}},"evidence":[{"@type":"CreativeWork","name":"the-decoder.com source article","url":"https://the-decoder.com/openai-shares-prompting-tips-for-gpt-6-astra-including-a-blocklist-of-slop-words","datePublished":"2026-09-05T13:31:04.000Z","provider":{"@type":"Organization","name":"the-decoder.com","url":"https://the-decoder.com/openai-shares-prompting-tips-for-gpt-6-astra-including-a-blocklist-of-slop-words"}},{"@type":"CreativeWork","name":"AIHot archive record","url":"https://aihot.virxact.com/items/cmtofayvo02uoroht238ucfjw","datePublished":"2026-09-05T13:31:04.000Z","provider":{"@type":"Organization","name":"AIHot","url":"https://aihot.virxact.com/items/cmtofayvo02uoroht238ucfjw"}}],"aggregationSource":"The Decoder：AI News（RSS）","originalPublisher":{"name":"the-decoder.com","url":"https://the-decoder.com/openai-shares-prompting-tips-for-gpt-6-astra-including-a-blocklist-of-slop-words"},"geoDeepAnswer":null,"article":{"id":"cmtofayvo02uoroht238ucfjw","slug":"cmtofayvo02uoroht238ucfjw","url":"https://www.aioga.com/news/cmtofayvo02uoroht238ucfjw/","title":"OpenAI 发布 GPT-6 Astra 提示词指南，含 slop 词屏蔽清单","title_en":"","summary":"OpenAI 在模型文档中说明 GPT-6 Astra 相比 GPT-5.6 Sol 更常提出澄清问题、对上下文更敏感，并给出让模型更主动、审计 AGENTS.md 等技能文件、控制写作风格、约束子智能体委派和测试规模的提示词建议。","source":"The Decoder：AI News（RSS）","sourceUrl":"https://the-decoder.com/openai-shares-prompting-tips-for-gpt-6-astra-including-a-blocklist-of-slop-words","aiHotUrl":"https://aihot.virxact.com/items/cmtofayvo02uoroht238ucfjw","publishedAt":"2026-09-05T13:31:04.000Z","category":"行业动态","score":72,"selected":true,"articleBody":["OpenAI's model documentation spells out where GPT-6 Astra tends toward unwanted behavior and how developers can work around it.","GPT-6 Astra asks clarifying questions more often than GPT-5.6 Sol instead of making assumptions on its own, according to OpenAI, making it a \"more effective collaborator.\" The trade-off is that the model sometimes stops where users expect it to keep going.","To push it toward more initiative, OpenAI recommends a prompt telling the model to infer the user's \"intent\" from context and show a \"bias towards action.\" Phrases like \"can you...,\" \"I want to...,\" or \"help me...\" should be treated as calls to act, not invitations for follow-up questions. Ad","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. Ad","The model should wait to ask for approval until it has already prepared a concrete, reviewable result. OpenAI prompts it this way: \"The user should be approving a concrete, reviewable result.\" Unsolicited warnings, disclaimers, or safety checklists based on hypothetical risks should be dropped from prompts.","GPT-6 Astra follows longer instructions better than its predecessors but is also more sensitive to context. Unclear or contradictory instructions in skill files like AGENTS.md can cause the model to block work or veer off unexpectedly. OpenAI recommends auditing all skill files and context documents the model can access and giving user instructions explicit priority. Ad","OpenAI also recommends a debugging prompt that forces the model to name the exact skill file and quote the specific instruction that caused it to pause or change direction. This helps developers trace unexpected behavior back to its source.","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. Ad","GPT-6 Astra tends to structure responses with lists, tables, and Markdown formatting, and reuses the same phrases across sessions. OpenAI has specific guidance on shaping the model's writing style. If you want prose, tell the model explicitly to write concise paragraphs using plain language and active voice. Ad","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.","A blocklist of typical AI phrases can help too. OpenAI calls them \"slop words.\" Made-up hyphenated compounds like \"exact-head checks\" or \"editorial-row layouts\" should also be avoided. The model should state what it's doing rather than listing what it won't do.","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.","For technical writing, OpenAI recommends keeping jargon to cases where it actually helps: \"Use plain language over jargon, and reference technical details only to the degree that it helps illustrate an idea or your work to the user.\"","The model can hand off work to sub-agents：https://developers.openai.com/api/docs/guides/responses-multi-agent running in parallel but does so less often than expected. Developers should spell out when and how much it should delegate, OpenAI says. Messages between agents can also contain grammar or spacing errors.","On coding tasks, GPT-6 Astra runs thorough tests before wrapping up. For small changes, that can mean test suites wildly out of proportion to the actual work. OpenAI recommends telling the model to rerun tests only when new failures or unresolved issues justify it.","More detailed versions of these prompts are available on the GPT-6 Astra model documentation page：https://developers.openai.com/api/docs/guides/latest-model. Developers who want to switch to GPT-6 Astra can use Codex with the OpenAI Docs skill：https://github.com/openai/skills/tree/main/skills/.curated/openai-docs to apply the recommended changes automatically: $openai-docs migrate this project to GPT-6 Astra","Stay in the loop on AI. Clear, useful, no fluff.","Follow The Decoder for AI news, background stories and expert analyses.","The Decoder：https://the-decoder.com/"],"articleImages":[{"sourceUrl":"https://the-decoder.com/wp-content/uploads/2026/09/openai_gpt_6_stars_astra.png","alt":"Image description","afterParagraph":0,"url":"/media/articles/cmtofayvo02uoroht238ucfjw/723f03f7bfa4374f.png"}],"mediaStatus":"ok","articleBodyZh":["OpenAI 的模型文档详细说明了 GPT-6 Astra 倾向出现的不良行为，以及开发者如何应对这些行为。","据 OpenAI 介绍，GPT-6 Astra 比 GPT-5.6 Sol 更频繁地提出澄清性问题，而不是自己做出假设，这使它成为“更高效的协作者”。但代价是模型有时会在用户期望它继续时停止。","为了让模型更主动，OpenAI 建议提示模型从上下文中推断用户的“意图”，并显示“行动倾向”。像“你能否……”、“我想要……”或“帮我……”这样的短语应被视为行动的指令，而不是跟进问题的邀请。","你应当从指示和迄今为止的对话上下文中推断用户的意图和任务范围。你的任务是展示行动倾向，并完成用户预期任务直到完成。如果用户表达了完成新工作或解决现有问题的意图，应持续工作直到实现用户的目标。除非操作明显具有破坏性或不可逆，否则应独立朝着用户目标工作（例如，创建独立的工作树/检出、解决合并冲突、执行只读操作、创建草稿 PR 等）。","模型应在准备好具体且可审查的结果后再请求批准。OpenAI 这样提示模型：“用户应当批准具体且可审查的结果。”不应在提示中包含基于假设风险的非必要警告、免责声明或安全清单。","GPT-6 Astra 对长指令的遵循比前代模型更好，但对上下文也更敏感。在 AGENTS.md 等技能文件中，指令不清晰或自相矛盾可能导致模型阻塞工作或意外偏离。OpenAI 建议审计模型可访问的所有技能文件和上下文文档，并明确优先执行用户指令。","OpenAI 还建议使用调试提示，强制模型指出导致其暂停或改变方向的具体技能文件，并引用具体指令。这帮助开发者追踪意外行为的来源。","如果某个技能会让你要求许可或确认、暂停、留下未完成的工作，或偏离用户的意图，请命名并链接你所阅读的确切 SKILL.md 文件，引用相关指令，并简要说明其适用性。区分明确的技能要求与你对指南的解释。 Ad","GPT-6 Astra 倾向于使用列表、表格和 Markdown 格式来组织回答，并在不同会话中重复使用相同的短语。OpenAI 对塑造模型的写作风格有具体指导。如果你想要散文，请明确告诉模型使用简明段落、通俗语言和主动语态来写。 Ad","默认情况下，使用清晰简明的段落，每段发展一个主要观点。仅在信息确实是平行的、顺序的或更容易比较时使用列表，并且除非层级关系无法用散文清楚表达，否则避免嵌套列表。使用简单直接的语言：熟悉的词汇、具体的例子和精准的动词。偏好主动语态和直接陈述。尽早明确主要观点，然后用读者所需的解释和细节进行扩展。让每个句子在前一句的基础上展开。发展重要论点，并提供足够的证据以便有用。","典型 AI 词语的黑名单也有帮助。OpenAI 称其为“废词”。应避免使用虚构的连字符复合词，如“exact-head checks”或“editorial-row layouts”。模型应说明自己正在做什么，而不是列出它不会做什么。","在结论中避免使用多余词语或短语，例如“结论：”“深入探讨”“促进”“使用/利用”“值得注意的是”“重要的是”“问题？答案”“这不是关于X，而是关于Y”“真的/确实”，或复合描述和连字符形容词。不要使用总结性的结尾语句，例如“简而言之：...”“最简单的思维模型是：...”。直接陈述预期动作。避免提及你不会做的内容、保持不变的内容或你将如何分离或分类结果。不要使用对比性措辞如“X，而不是Y”或“X——不是Y”，这会引入用户未请求的备选方案。避免捏造的复合术语如“exact-head checks”和“editorial-row layouts”、模糊限定词和固定过渡语；使用简单的动词和介词直接表达实际关系。","对于技术写作，OpenAI建议仅在确实有帮助的情况下使用术语：\"优先使用通俗语言而非术语，并仅在有助于向用户说明一个想法或你的工作时引用技术细节。\"","模型可以将工作交给子代理执行：https://developers.openai.com/api/docs/guides/responses-multi-agent 可以并行运行，但实际使用频率低于预期。OpenAI表示，开发者应明确说明什么时候以及委托多少工作。代理之间的消息可能存在语法或空格错误。","在编码任务上，GPT-6 Astra 在完成前会进行全面测试。对于小修改，这可能意味着测试套件远远超过实际工作所需。OpenAI建议仅在出现新的失败或未解决问题时，才让模型重新运行测试。","这些提示的更详细版本可在 GPT-6 Astra 模型文档页面找到：https://developers.openai.com/api/docs/guides/latest-model。希望切换到 GPT-6 Astra 的开发者可以使用带有 OpenAI Docs 技能的 Codex：https://github.com/openai/skills/tree/main/skills/.curated/openai-docs 自动应用推荐更改：$openai-docs migrate this project to GPT-6 Astra","保持掌握 AI 动向。清楚、有用，无废话。","关注 The Decoder 获取 AI 新闻、背景故事和专家分析。","解码器：https://the-decoder.com/"],"translationStatus":"translated","bodyOrigin":"source-page","editorial":{"summary":"OpenAI 模型文档介绍了 GPT-6 Astra 的行为特点与提示词建议。文档称，该模型比 GPT-5.6 Sol 更常提出澄清问题，也更重视上下文；同时提供了推动模型主动执行、审计技能文件、控制写作风格、约束子智能体委派和测试规模的方法。","background":"根据来源材料，GPT-6 Astra 有时会在用户期待其继续推进时停止。OpenAI 建议提示模型从对话上下文推断意图，并对行动保持倾向；还建议审查 AGENTS.md 等技能文件，明确用户指令优先级，并通过引用具体文件和指令定位异常行为来源。","viewpoint":"Aioga 判断：这份指南的重点不只是提示词措辞，而是把模型行为调整与上下文治理、技能文件审计和结果复核联系起来。对于使用复杂指令链的团队，提示词和可访问文档都值得纳入排查范围。","implications":"可能影响：开发者需要关注更频繁的澄清行为与过早停止之间的取舍。建议在提示词中明确行动边界和任务意图，并审查可能影响执行方向的技能文件；这些做法可能改善可追溯性，但不代表能够消除所有偏离。","nextStep":"后续观察：应继续关注 OpenAI 模型文档是否补充 GPT-6 Astra 在不同任务中的行为说明，以及相关提示词建议是否扩展到更多上下文管理、委派控制和测试实践。","evidenceRefs":["title","summary","articleBody","source"],"status":"published","aiGenerated":true,"autoApproved":true,"generatedBy":"aioga-editorial:gpt-5.6-sol","reviewedBy":"aioga-editorial-review:gpt-5.6-sol","generatedAt":"2026-09-05T13:44:26.633Z","sourceHash":"33fce9eccd2e49de","review":{"approved":true,"groundedness":95,"clarity":92,"duplicationRisk":10,"blockingIssues":[],"notes":["候选内容与来源材料整体一致，未发现事实错误、无来源断言或将观点冒充事实的情况。","viewpoint 已明确标注为判断，implications 使用“可能”等限定性表述，未将推论绝对化。","候选内容未提及来源标题中的“slop 词屏蔽清单”，但这属于信息取舍，不构成事实错误或阻断问题。"]},"validation":{"passed":true,"mode":"ai-auto","revisions":0,"checks":["schema","length","source-attribution","editorial-labels","inference-boundary","low-source-overlap","no-html","independent-ai-review"]}},"tags":["行业动态","The Decoder：AI News（RSS）"],"translations":{"zh-CN":{"title":"OpenAI 发布 GPT-6 Astra 提示词指南，含 slop 词屏蔽清单","summary":"OpenAI 在模型文档中说明 GPT-6 Astra 相比 GPT-5.6 Sol 更常提出澄清问题、对上下文更敏感，并给出让模型更主动、审计 AGENTS.md 等技能文件、控制写作风格、约束子智能体委派和测试规模的提示词建议。","category":"行业动态","source":"the-decoder.com","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI 发布 GPT-6 Astra 提示词指南，含 slop 词屏蔽清单 - Aioga AI资讯","description":"OpenAI 在模型文档中说明 GPT-6 Astra 相比 GPT-5.6 Sol 更常提出澄清问题、对上下文更敏感，并给出让模型更主动、审计 AGENTS.md 等技能文件、控制写作风格、约束子智能体委派和测试规模的提示词建议。","url":"https://www.aioga.com/news/cmtofayvo02uoroht238ucfjw/","articleBody":["OpenAI 的模型文档详细说明了 GPT-6 Astra 倾向出现的不良行为，以及开发者如何应对这些行为。","据 OpenAI 介绍，GPT-6 Astra 比 GPT-5.6 Sol 更频繁地提出澄清性问题，而不是自己做出假设，这使它成为“更高效的协作者”。但代价是模型有时会在用户期望它继续时停止。","为了让模型更主动，OpenAI 建议提示模型从上下文中推断用户的“意图”，并显示“行动倾向”。像“你能否……”、“我想要……”或“帮我……”这样的短语应被视为行动的指令，而不是跟进问题的邀请。","你应当从指示和迄今为止的对话上下文中推断用户的意图和任务范围。你的任务是展示行动倾向，并完成用户预期任务直到完成。如果用户表达了完成新工作或解决现有问题的意图，应持续工作直到实现用户的目标。除非操作明显具有破坏性或不可逆，否则应独立朝着用户目标工作（例如，创建独立的工作树/检出、解决合并冲突、执行只读操作、创建草稿 PR 等）。","模型应在准备好具体且可审查的结果后再请求批准。OpenAI 这样提示模型：“用户应当批准具体且可审查的结果。”不应在提示中包含基于假设风险的非必要警告、免责声明或安全清单。","GPT-6 Astra 对长指令的遵循比前代模型更好，但对上下文也更敏感。在 AGENTS.md 等技能文件中，指令不清晰或自相矛盾可能导致模型阻塞工作或意外偏离。OpenAI 建议审计模型可访问的所有技能文件和上下文文档，并明确优先执行用户指令。","OpenAI 还建议使用调试提示，强制模型指出导致其暂停或改变方向的具体技能文件，并引用具体指令。这帮助开发者追踪意外行为的来源。","如果某个技能会让你要求许可或确认、暂停、留下未完成的工作，或偏离用户的意图，请命名并链接你所阅读的确切 SKILL.md 文件，引用相关指令，并简要说明其适用性。区分明确的技能要求与你对指南的解释。 Ad","GPT-6 Astra 倾向于使用列表、表格和 Markdown 格式来组织回答，并在不同会话中重复使用相同的短语。OpenAI 对塑造模型的写作风格有具体指导。如果你想要散文，请明确告诉模型使用简明段落、通俗语言和主动语态来写。 Ad","默认情况下，使用清晰简明的段落，每段发展一个主要观点。仅在信息确实是平行的、顺序的或更容易比较时使用列表，并且除非层级关系无法用散文清楚表达，否则避免嵌套列表。使用简单直接的语言：熟悉的词汇、具体的例子和精准的动词。偏好主动语态和直接陈述。尽早明确主要观点，然后用读者所需的解释和细节进行扩展。让每个句子在前一句的基础上展开。发展重要论点，并提供足够的证据以便有用。","典型 AI 词语的黑名单也有帮助。OpenAI 称其为“废词”。应避免使用虚构的连字符复合词，如“exact-head checks”或“editorial-row layouts”。模型应说明自己正在做什么，而不是列出它不会做什么。","在结论中避免使用多余词语或短语，例如“结论：”“深入探讨”“促进”“使用/利用”“值得注意的是”“重要的是”“问题？答案”“这不是关于X，而是关于Y”“真的/确实”，或复合描述和连字符形容词。不要使用总结性的结尾语句，例如“简而言之：...”“最简单的思维模型是：...”。直接陈述预期动作。避免提及你不会做的内容、保持不变的内容或你将如何分离或分类结果。不要使用对比性措辞如“X，而不是Y”或“X——不是Y”，这会引入用户未请求的备选方案。避免捏造的复合术语如“exact-head checks”和“editorial-row layouts”、模糊限定词和固定过渡语；使用简单的动词和介词直接表达实际关系。","对于技术写作，OpenAI建议仅在确实有帮助的情况下使用术语：\"优先使用通俗语言而非术语，并仅在有助于向用户说明一个想法或你的工作时引用技术细节。\"","模型可以将工作交给子代理执行：https://developers.openai.com/api/docs/guides/responses-multi-agent 可以并行运行，但实际使用频率低于预期。OpenAI表示，开发者应明确说明什么时候以及委托多少工作。代理之间的消息可能存在语法或空格错误。","在编码任务上，GPT-6 Astra 在完成前会进行全面测试。对于小修改，这可能意味着测试套件远远超过实际工作所需。OpenAI建议仅在出现新的失败或未解决问题时，才让模型重新运行测试。","这些提示的更详细版本可在 GPT-6 Astra 模型文档页面找到：https://developers.openai.com/api/docs/guides/latest-model。希望切换到 GPT-6 Astra 的开发者可以使用带有 OpenAI Docs 技能的 Codex：https://github.com/openai/skills/tree/main/skills/.curated/openai-docs 自动应用推荐更改：$openai-docs migrate this project to GPT-6 Astra","保持掌握 AI 动向。清楚、有用，无废话。","关注 The Decoder 获取 AI 新闻、背景故事和专家分析。","解码器：https://the-decoder.com/"]},"en":{"title":"OpenAI Releases GPT-6 Astra Prompt Guide, Including Slop Word Blocklist","summary":"OpenAI states in the model documentation that compared to GPT-5.6 Sol, GPT-6 Astra more frequently asks clarifying questions, is more context-aware, and provides prompt suggestions to make the model more proactive, audit skills files such as AGENTS.md, control writing style, constrain sub-agent delegation, and test scale.","category":"Industry","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI Releases GPT-6 Astra Prompt Guide, Including Slop Word Blocklist - Aioga AI News","description":"OpenAI states in the model documentation that compared to GPT-5.6 Sol, GPT-6 Astra more frequently asks clarifying questions, is more context-aware, and provides prompt suggestions...","url":"https://www.aioga.com/en/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:06.101Z"},"ja":{"title":"OpenAI、GPT-6 Astraプロンプトガイドを公開、slopワードブロックリストを含む","summary":"OpenAIはモデルドキュメントで、GPT-6 AstraはGPT-5.6 Solに比べてより多く確認質問を行い、文脈に敏感であることを説明し、モデルをより積極的に動かすためのプロンプト、AGENTS.mdなどのスキルファイルを監査する、文章スタイルを制御する、サブエージェントの委任とテスト範囲を制約するためのプロンプト提案を示した。","category":"業界動向","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI、GPT-6 Astraプロンプトガイドを公開、slopワードブロックリストを含む - Aioga AIニュース","description":"OpenAIはモデルドキュメントで、GPT-6 AstraはGPT-5.6 Solに比べてより多く確認質問を行い、文脈に敏感であることを説明し、モデルをより積極的に動かすためのプロンプト、AGENTS.mdなどのスキルファイルを監査する、文章スタイルを制御する、サブエージェントの委任とテスト範囲を制約するためのプロンプト提案を示した。","url":"https://www.aioga.com/ja/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:08.144Z"},"ko":{"title":"OpenAI, GPT-6 Astra 프롬프트 가이드 공개, slop 단어 차단 목록 포함","summary":"OpenAI는 모델 문서에서 GPT-6 Astra가 GPT-5.6 Sol에 비해 더 자주 확인 질문을 제시하고, 문맥에 더 민감하며, 모델을 더 적극적으로 만들고, AGENTS.md 등 기술 문서를 감사하며, 글쓰기 스타일을 제어하고, 하위 에이전트 위임을 제약하며, 테스트 규모를 조정하는 프롬프트 권장사항을 제공한다고 설명했다.","category":"업계 동향","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI, GPT-6 Astra 프롬프트 가이드 공개, slop 단어 차단 목록 포함 - Aioga AI 뉴스","description":"OpenAI는 모델 문서에서 GPT-6 Astra가 GPT-5.6 Sol에 비해 더 자주 확인 질문을 제시하고, 문맥에 더 민감하며, 모델을 더 적극적으로 만들고, AGENTS.md 등 기술 문서를 감사하며, 글쓰기 스타일을 제어하고, 하위 에이전트 위임을 제약하며, 테스트 규모를 조정하는 프롬프트 권장사항을 제공한다고...","url":"https://www.aioga.com/ko/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:12.249Z"},"es":{"title":"OpenAI publica la guía de prompts de GPT-6 Astra, incluyendo la lista de palabras slop bloqueadas","summary":"OpenAI indica en la documentación del modelo que GPT-6 Astra, en comparación con GPT-5.6 Sol, plantea con mayor frecuencia preguntas de aclaración, es más sensible al contexto y proporciona sugerencias de prompts para que el modelo sea más proactivo, audite archivos de habilidades como AGENTS.md, controle el estilo de escritura, restrinja la delegación de subagentes y pruebe la escala.","category":"Industria","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI publica la guía de prompts de GPT-6 Astra, incluyendo la lista de palabras slop bloqueadas - Aioga Noticias de IA","description":"OpenAI indica en la documentación del modelo que GPT-6 Astra, en comparación con GPT-5.6 Sol, plantea con mayor frecuencia preguntas de aclaración, es más sensible al contexto y pr...","url":"https://www.aioga.com/es/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:12.298Z"},"fr":{"title":"OpenAI publie le guide des prompts GPT-6 Astra, incluant une liste de mots à filtrer \"slop\"","summary":"OpenAI indique dans la documentation du modèle que par rapport à GPT-5.6 Sol, GPT-6 Astra pose plus souvent des questions de clarification, est plus sensible au contexte, et donne des suggestions de prompts permettant de rendre le modèle plus proactif, d'auditer des fichiers de compétences comme AGENTS.md, de contrôler le style d'écriture, de restreindre la délégation aux sous-agents et de tester l'échelle.","category":"Industrie","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI publie le guide des prompts GPT-6 Astra, incluant une liste de mots à filtrer \"slop\" - Aioga Actualités IA","description":"OpenAI indique dans la documentation du modèle que par rapport à GPT-5.6 Sol, GPT-6 Astra pose plus souvent des questions de clarification, est plus sensible au contexte, et donne...","url":"https://www.aioga.com/fr/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:16.175Z"},"de":{"title":"OpenAI veröffentlicht GPT-6 Astra Prompt-Leitfaden, inklusive Liste blockierter Slop-Wörter","summary":"OpenAI erklärt in den Modelldokumentationen, dass GPT-6 Astra im Vergleich zu GPT-5.6 Sol häufiger Klärungsfragen stellt, kontextsensitiver ist und Prompt-Vorschläge gibt, um das Modell proaktiver zu machen, AGENTS.md und andere Skill-Dateien zu prüfen, Schreibstil zu steuern, Delegation an Sub-Intelligenzen zu begrenzen und die Testskala zu kontrollieren.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI veröffentlicht GPT-6 Astra Prompt-Leitfaden, inklusive Liste blockierter Slop-Wörter - Aioga KI-News","description":"OpenAI erklärt in den Modelldokumentationen, dass GPT-6 Astra im Vergleich zu GPT-5.6 Sol häufiger Klärungsfragen stellt, kontextsensitiver ist und Prompt-Vorschläge gibt, um das M...","url":"https://www.aioga.com/de/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:16.433Z"},"pt-BR":{"title":"OpenAI lança guia de prompts do GPT-6 Astra, incluindo lista de palavras slop a serem bloqueadas","summary":"A OpenAI explicou nos documentos do modelo que o GPT-6 Astra, em comparação com o GPT-5.6 Sol, faz perguntas de esclarecimento com mais frequência, é mais sensível ao contexto e fornece sugestões de prompts para tornar o modelo mais proativo, auditar arquivos de habilidades como AGENTS.md, controlar o estilo de escrita, restringir delegação de subagentes e a escala de testes.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI lança guia de prompts do GPT-6 Astra, incluindo lista de palavras slop a serem bloqueadas - Aioga Notícias de IA","description":"A OpenAI explicou nos documentos do modelo que o GPT-6 Astra, em comparação com o GPT-5.6 Sol, faz perguntas de esclarecimento com mais frequência, é mais sensível ao contexto e fo...","url":"https://www.aioga.com/pt-BR/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:20.389Z"},"ru":{"title":"OpenAI выпустила руководство по подсказкам GPT-6 Astra, включая список заблокированных слов slop","summary":"OpenAI в документации к модели указала, что GPT-6 Astra по сравнению с GPT-5.6 Sol чаще задаёт уточняющие вопросы, более чувствителен к контексту и предоставляет рекомендации по подсказкам, которые позволяют модели действовать активнее, проверять файлы навыков, такие как AGENTS.md, контролировать стиль письма, ограничивать делегирование подагентов и масштаб тестирования.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI выпустила руководство по подсказкам GPT-6 Astra, включая список заблокированных слов slop - Aioga Новости ИИ","description":"OpenAI в документации к модели указала, что GPT-6 Astra по сравнению с GPT-5.6 Sol чаще задаёт уточняющие вопросы, более чувствителен к контексту и предоставляет рекомендации по по...","url":"https://www.aioga.com/ru/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:21.569Z"},"ar":{"title":"أعلنت OpenAI عن دليل الإرشادات لنماذج GPT-6 Astra، بما في ذلك قائمة كلمات slop الممنوعة","summary":"أوضحت OpenAI في وثائق النماذج أن GPT-6 Astra بالمقارنة مع GPT-5.6 Sol يطرح أسئلة توضيحية بشكل أكثر شيوعًا، ويكون أكثر حساسية للسياق، ويقدم اقتراحات لجعل النموذج أكثر نشاطًا، ومراجعة ملفات المهارات مثل AGENTS.md، والتحكم في أسلوب الكتابة، وتقييد تفويض الوكلاء الفرعيين واختبار النطاق.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"أعلنت OpenAI عن دليل الإرشادات لنماذج GPT-6 Astra، بما في ذلك قائمة كلمات slop الممنوعة - Aioga أخبار الذكاء الاصطناعي","description":"أوضحت OpenAI في وثائق النماذج أن GPT-6 Astra بالمقارنة مع GPT-5.6 Sol يطرح أسئلة توضيحية بشكل أكثر شيوعًا، ويكون أكثر حساسية للسياق، ويقدم اقتراحات لجعل النموذج أكثر نشاطًا، ومراجع...","url":"https://www.aioga.com/ar/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:25.628Z"},"hi":{"title":"OpenAI ने GPT-6 Astra प्रॉम्प्ट गाइड जारी किया, जिसमें slop शब्द ब्लॉक सूची शामिल है","summary":"OpenAI ने मॉडल दस्तावेज़ में बताया कि GPT-6 Astra की तुलना में GPT-5.6 Sol की तुलना में अधिक बार स्पष्टता सवाल पूछता है, संदर्भ के प्रति अधिक संवेदनशील है, और मॉडलों को अधिक सक्रिय बनाने, AGENTS.md जैसी कौशल फ़ाइलों का ऑडिट करने, लेखन शैली को नियंत्रित करने, उप-बुद्धिमान एजेंट के डेलीगेशन को प्रतिबंधित करने और परीक्षण पैमाने के लिए प्रॉम्प्ट सुझाव देने का तरीका बताता है।","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI ने GPT-6 Astra प्रॉम्प्ट गाइड जारी किया, जिसमें slop शब्द ब्लॉक सूची शामिल है - Aioga AI समाचार","description":"OpenAI ने मॉडल दस्तावेज़ में बताया कि GPT-6 Astra की तुलना में GPT-5.6 Sol की तुलना में अधिक बार स्पष्टता सवाल पूछता है, संदर्भ के प्रति अधिक संवेदनशील है, और मॉडलों को अधिक सक्रिय...","url":"https://www.aioga.com/hi/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:26.670Z"},"it":{"title":"OpenAI rilascia la guida ai prompt di GPT-6 Astra, inclusa la lista di parole slop da bloccare","summary":"OpenAI spiega nella documentazione del modello che GPT-6 Astra, rispetto a GPT-5.6 Sol, è più propenso a porre domande di chiarimento, è più sensibile al contesto e fornisce suggerimenti sui prompt per rendere il modello più proattivo, esaminare file di competenze come AGENTS.md, controllare lo stile di scrittura, vincolare la delega ai sotto-agenti e testare la scala.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI rilascia la guida ai prompt di GPT-6 Astra, inclusa la lista di parole slop da bloccare - Aioga Notizie IA","description":"OpenAI spiega nella documentazione del modello che GPT-6 Astra, rispetto a GPT-5.6 Sol, è più propenso a porre domande di chiarimento, è più sensibile al contesto e fornisce sugger...","url":"https://www.aioga.com/it/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:32.059Z"},"nl":{"title":"OpenAI publiceert GPT-6 Astra promptgids, inclusief lijst met geblokkeerde 'slop'-woorden","summary":"OpenAI legt in de modeldocumentatie uit dat GPT-6 Astra in vergelijking met GPT-5.6 Sol vaker om verduidelijking vraagt, gevoeliger is voor context en suggesties biedt voor prompts waarmee het model proactiever kan zijn, AGENTS.md en andere vaardigheidsbestanden kan controleren, schrijfstijl kan aanpassen, delegatie van subagenten kan beperken en testschaal kan beheren.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI publiceert GPT-6 Astra promptgids, inclusief lijst met geblokkeerde 'slop'-woorden - Aioga AI-nieuws","description":"OpenAI legt in de modeldocumentatie uit dat GPT-6 Astra in vergelijking met GPT-5.6 Sol vaker om verduidelijking vraagt, gevoeliger is voor context en suggesties biedt voor prompts...","url":"https://www.aioga.com/nl/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:30.697Z"},"tr":{"title":"OpenAI, GPT-6 Astra için İpucu Kılavuzu Yayınladı, slop kelime engelleme listesi dahil","summary":"OpenAI, model belgelerinde GPT-6 Astra'nın GPT-5.6 Sol'a kıyasla daha sık açıklama soruları sorduğunu, bağlama daha duyarlı olduğunu ve modeli daha proaktif hale getirme, AGENTS.md gibi beceri dosyalarını denetleme, yazı stilini kontrol etme, alt zekâ birimi devretme ve test ölçeğini sınırlama için ipucu önerileri verdiğini belirtti.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI, GPT-6 Astra için İpucu Kılavuzu Yayınladı, slop kelime engelleme listesi dahil - Aioga AI Haberleri","description":"OpenAI, model belgelerinde GPT-6 Astra'nın GPT-5.6 Sol'a kıyasla daha sık açıklama soruları sorduğunu, bağlama daha duyarlı olduğunu ve modeli daha proaktif hale getirme, AGENTS.md...","url":"https://www.aioga.com/tr/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:36.658Z"},"vi":{"title":"OpenAI phát hành hướng dẫn prompt GPT-6 Astra, bao gồm danh sách từ ngăn chặn slop","summary":"OpenAI trong tài liệu mô hình cho biết GPT-6 Astra so với GPT-5.6 Sol thường đưa ra các câu hỏi làm rõ hơn, nhạy cảm hơn với ngữ cảnh, đồng thời cung cấp các gợi ý prompt giúp mô hình chủ động hơn, kiểm tra các tệp kỹ năng như AGENTS.md, kiểm soát phong cách viết, ràng buộc việc ủy quyền cho các tiểu trí tuệ và thử nghiệm quy mô.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI phát hành hướng dẫn prompt GPT-6 Astra, bao gồm danh sách từ ngăn chặn slop - Tin tức AI Aioga","description":"OpenAI trong tài liệu mô hình cho biết GPT-6 Astra so với GPT-5.6 Sol thường đưa ra các câu hỏi làm rõ hơn, nhạy cảm hơn với ngữ cảnh, đồng thời cung cấp các gợi ý prompt giúp mô h...","url":"https://www.aioga.com/vi/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:35.986Z"},"id":{"title":"OpenAI merilis panduan prompt GPT-6 Astra, termasuk daftar blokir kata slop","summary":"OpenAI menjelaskan dalam dokumentasi model bahwa dibandingkan dengan GPT-5.6 Sol, GPT-6 Astra lebih sering mengajukan pertanyaan klarifikasi, lebih sensitif terhadap konteks, serta memberikan saran prompt untuk membuat model lebih proaktif, mengaudit file keterampilan seperti AGENTS.md, mengontrol gaya penulisan, membatasi delegasi sub-agent, dan skala pengujian.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI merilis panduan prompt GPT-6 Astra, termasuk daftar blokir kata slop - Berita AI Aioga","description":"OpenAI menjelaskan dalam dokumentasi model bahwa dibandingkan dengan GPT-5.6 Sol, GPT-6 Astra lebih sering mengajukan pertanyaan klarifikasi, lebih sensitif terhadap konteks, serta...","url":"https://www.aioga.com/id/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:40.826Z"},"th":{"title":"OpenAI เผยแพร่คู่มือคำสั่ง GPT-6 Astra พร้อมรายการคำ slop ที่ต้องบล็อก","summary":"OpenAI อธิบายในเอกสารโมเดลว่า GPT-6 Astra เมื่อเทียบกับ GPT-5.6 Sol มักถามคำถามเพื่อชี้แจงมากขึ้น มีความไวต่อบริบทมากขึ้น และให้คำแนะนำเกี่ยวกับคำสั่งเพื่อทำให้โมเดลมีความรุกเร้ามากขึ้น ตรวจสอบไฟล์ทักษะ AGENTS.md ควบคุมสไตล์การเขียน จำกัดการมอบหมายของ sub-agent และทดสอบขนาดของโมเดล","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI เผยแพร่คู่มือคำสั่ง GPT-6 Astra พร้อมรายการคำ slop ที่ต้องบล็อก - ข่าว AI Aioga","description":"OpenAI อธิบายในเอกสารโมเดลว่า GPT-6 Astra เมื่อเทียบกับ GPT-5.6 Sol มักถามคำถามเพื่อชี้แจงมากขึ้น มีความไวต่อบริบทมากขึ้น และให้คำแนะนำเกี่ยวกับคำสั่งเพื่อทำให้โมเดลมีความรุกเร้ามา...","url":"https://www.aioga.com/th/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:42.461Z"},"pl":{"title":"OpenAI wydaje przewodnik po promptach GPT-6 Astra, zawierający listę blokad słów slop","summary":"OpenAI w dokumentacji modelu wyjaśnia, że GPT-6 Astra w porównaniu z GPT-5.6 Sol częściej zadaje pytania wyjaśniające, jest bardziej wrażliwy na kontekst oraz podaje sugestie promptów umożliwiające modelowi większą proaktywność, audytowanie plików umiejętności takich jak AGENTS.md, kontrolowanie stylu pisania, ograniczanie delegowania podumysłów i testowanie skali.","category":"行业动态","source":"The Decoder：AI News（RSS）","aggregationSource":"The Decoder：AI News（RSS）","pageTitle":"OpenAI wydaje przewodnik po promptach GPT-6 Astra, zawierający listę blokad słów slop - Aioga Wiadomości AI","description":"OpenAI w dokumentacji modelu wyjaśnia, że GPT-6 Astra w porównaniu z GPT-5.6 Sol częściej zadaje pytania wyjaśniające, jest bardziej wrażliwy na kontekst oraz podaje sugestie promp...","url":"https://www.aioga.com/pl/news/cmtofayvo02uoroht238ucfjw/","contentTranslated":true,"sourceHash":"c547eb541a0fbef3","translatedAt":"2026-09-05T13:41:47.003Z"}},"evidenceTier":"verified-news","reviewStatus":"editorial-selected","indexable":true,"editorialCover":""}}