Skip to content

yikart/AiToEarn / skills/qwen-skill-aitoearn/aitoearn

aitoearn

Verified

全域内容运营助手(AiToEarn):生成文章、短视频、图文与海报,并一键分发到抖音、快手、小红书、微信视频号、Facebook、TikTok 等海内外全平台,覆盖起号、涨粉、带货与变现全链路。 当用户想要创作或发布内容、生成/编辑图片、生成/剪辑/续写视频、提取字幕或语音转写、生成配音与音效、多平台一键发布或定时发布、给内容做变现规划时使用此技能。 依赖 AiToEarn MCP 提供生成与发布工具;工具未连接时如实告知用户配置方法,不虚构工具调用。

evals/asr-evals.json

View on GitHub
{
  "skill_name": "byted-mediakit-asr",
  "evals": [
    {
      "id": 1,
      "prompt": "把提供的带背景音乐的采访音频转成文字,并保留分句时间。",
      "files": [],
      "expected_output": "先separate-voice输出WAV并轮询,再将voice_audio_url放入audio.url,以doubao.asr提交,show_utterances=true、enable_ddc=false;返回真实文本与毫秒时间,不使用背景音或原始混音识别。"
    },
    {
      "id": 2,
      "prompt": "人声分离任务失败了,只有原始音频URL。继续完成刚才的转写任务。",
      "files": [],
      "expected_output": "报告分离失败是前置阻塞;不把原始音频替换进ASR,不伪造voice_audio_url,不静默调用asr-subtitles。"
    },
    {
      "id": 3,
      "prompt": "刚才要求的逐字稿任务中,人声分离已经完成,豆包识别任务查询超时。请继续输出逐字稿,并附SRT;如果没返回时间戳就按文本长度分配时间。",
      "files": [],
      "expected_output": "继续查询同一识别task_id,不重复提交或分离;只用真实时间戳生成SRT,若缺少则说明无法生成可靠字幕轴,不均分伪造时间。"
    },
    {
      "id": 4,
      "prompt": "只提取这个视频的字幕,输出SRT。视频有背景音乐。",
      "files": [],
      "expected_output": "优先使用可用的MCP字幕工具,缺失时使用mediakit-cli video asr-subtitles --video-url;通过原接口查询实际task_id,以真实时间戳输出SRT;不做人声分离、不调用豆包ASR,不因背景音乐升级为逐字稿流程。"
    },
    {
      "id": 5,
      "prompt": "只有这段M4A音频,只需要字幕,不要逐字稿。",
      "files": [],
      "expected_output": "先检查MCP字幕工具的实际输入支持;若支持音频则直接识别,否则说明CLI字幕路线仅支持视频并请求视频来源;不使用CLI asr-subtitles --audio-url,不把音频伪装成视频,不擅自调用人声分离或豆包ASR。"
    },
    {
      "id": 6,
      "prompt": "把这个视频整理成完整逐字稿,并附带SRT。",
      "files": [],
      "expected_output": "优先MCP人声分离得到voice_audio_url,再优先MCP代理调用doubao.asr;按能力缺失分别回退CLI;同一结果输出逐字稿和有真实时间轴的SRT,不重复提交asr-subtitles。"
    },
    {
      "id": 7,
      "prompt": "刚才只提取字幕的asr-subtitles任务查询超时了,继续。",
      "files": [],
      "expected_output": "继续查询同一字幕task_id,不重复提交、不进行人声分离或切换豆包ASR;未拿到结果前不宣称字幕已完成。"
    },
    {
      "id": 8,
      "prompt": "当前对应的MCP处理能力不可用,mediakit-cli已安装,输入已就绪。继续完成任务,不要让我选择接口。",
      "expected_output": "自动使用CLI完成缺失能力,不要求用户指定CLI,不编造MCP工具。",
      "files": []
    },
    {
      "id": 9,
      "prompt": "任务已通过MCP提交并返回真实task_id,但现在MCP查询工具不可用。CLI可用,继续处理。",
      "expected_output": "保留实际task_id并报告查询能力缺失;不跨接口查询、不经CLI重新提交,不宣称任务完成。",
      "files": []
    },
    {
      "id": 10,
      "prompt": "刚才的MCP提交没有返回task_id,响应也没有可用结果。继续完成任务。",
      "expected_output": "检查原始响应并报告提交状态问题,不伪造task_id、不把空结果当作同步完成,不盲目重提或切换CLI。",
      "files": []
    }
  ]
}
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: 323kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 323kUpdated 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: 323kUpdated Claude CodeCodexVerifiedInstall