Skip to content

ahang008/ah-skills / skills/ah-gemini-script-polisher

ah-gemini-script-polisher

Verified

使用 Gemini 润色或创作中文口播稿,保留作者事实、观点和表达。用于用户要求 Gemini 写口播稿、让 Flash 润色、把 Codex 上下文交给 Gemini 写作、调用口播写作桥,或讨论此桥的架构。普通口播写作未指定 Gemini、通用 Gemini 研究和应用代码架构不触发。

FileKindSize
README.mdmarkdown3.9 KB
SKILL.mdskill md6.0 KB
agents/openai.yamlconfig297 B
references/01-context-packet.mdmarkdown2.3 KB
references/02-browser-bridge.mdmarkdown4.6 KB
scripts/check_response.pyscript3.1 KB
tests/test_response.pyscript2.3 KB

references/01-context-packet.md

View on GitHub
# 写作包模板

只保留本次稿件需要的信息。事实字段注明依据;来源可写“当次用户输入”等短标签,不把本地绝对路径、历史对话或敏感资料粘贴到 Gemini。

## 输入模板

```text
任务: {{task_id}}
版本: {{revision}}
模式: 润色已有稿 / 根据提纲成稿
主题: {{主题}}
受众和目的: {{受众}};{{观众看完应获得什么}}
长度要求: {{字数范围或时长,未指定时不捏造要求}}
作者身份: {{与本稿有关的身份,例如产品经理出身的商人}}

原稿或观点提纲:
{{完整原稿;或按原始表达整理的提纲和素材}}

已确认事实和禁止扩大边界:
- F1: {{事实、单位、条件、依据}}
- F2: {{事实、单位、条件、依据}}

必须保留的原话或观点:
{{真正需要锁定的内容,没有则写无}}

风格参考:
{{作者当前表达或已确认的短片段;只提供有用的例子}}

本次写作要求:
写成适合直接念的中文口播,改善措辞、节奏和衔接。允许调整句序和口语表达,保留作者立场。正文不要写标题、分镜、时间码、解释、协议或写作评分。遵守本次长度范围。不要补造经历、数字、结果或商业承诺;关键事实缺失时返回具体缺口,不编成定稿。

正常完成时严格返回:
SCRIPT_BEGIN task_id={{task_id}} revision={{revision}}
{{完整口播正文}}
SCRIPT_END task_id={{task_id}} revision={{revision}}
修改说明: {{简短列出实际修改,没有事实变更就写未新增事实}}
REPLY_DONE task_id={{task_id}} revision={{revision}}

事实不足以成稿时返回:
NEEDS_INPUT task_id={{task_id}} revision={{revision}}
{{真正阻止成稿的具体问题}}
REPLY_DONE task_id={{task_id}} revision={{revision}}
```

## 定点修订

保留原任务 ID,版本加 1。发送上版问题句、对应 F 编号和明确修改要求;请求重新返回完整正文和新版本标识,不仅返回替换片段。既有风格和其他已通过内容不需重改。

## 人工审核的边界

- 主体、量词、数字、时间范围、语气强弱和因果都属于事实与原意检查,不仅检查数字是否一致。
- 没有事实缺陷、原意偏移或明显拗口,就保留 Gemini 原文。
- 时长只能按阅读速度估算,未经实际口读不声称实测时长。
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall