OpenClawSkills
GitHub
节点和媒体 • TutorialHeader.readTime

媒体理解

ReferenceNodesMediaUnderstandingPage.header.description

ReferenceNodesMediaUnderstandingPage.intro.p1

Tutorial.step

目标

ReferenceNodesMediaUnderstandingPage.steps.01.item1

ReferenceNodesMediaUnderstandingPage.steps.01.item2

ReferenceNodesMediaUnderstandingPage.steps.01.item3

ReferenceNodesMediaUnderstandingPage.steps.01.item4

Tutorial.step

ReferenceNodesMediaUnderstandingPage.steps.02.title

ReferenceNodesMediaUnderstandingPage.steps.02.item1

ReferenceNodesMediaUnderstandingPage.steps.02.item2

ReferenceNodesMediaUnderstandingPage.steps.02.item3

ReferenceNodesMediaUnderstandingPage.steps.02.item4

5. 成功時：

ReferenceNodesMediaUnderstandingPage.steps.02.item5a

ReferenceNodesMediaUnderstandingPage.steps.02.item5b

ReferenceNodesMediaUnderstandingPage.steps.02.item5c

ReferenceNodesMediaUnderstandingPage.steps.02.item5d

ReferenceNodesMediaUnderstandingPage.steps.02.item6

Tutorial.step

设置的概览

ReferenceNodesMediaUnderstandingPage.steps.03.p1

ReferenceNodesMediaUnderstandingPage.steps.03.item1

- tools.media.image / tools.media.audio / tools.media.video：

- 默认（prompt、maxChars、maxBytes、timeoutSeconds、language）

- 提供商覆盖（baseUrl、headers、providerOptions）

ReferenceNodesMediaUnderstandingPage.steps.03.item2c

ReferenceNodesMediaUnderstandingPage.steps.03.item2d

- attachments 策略（mode、maxAttachments、prefer）

ReferenceNodesMediaUnderstandingPage.steps.03.item4

- tools.media.concurrency：最大同時機能运行数（默认 2）。

Json5
{
  tools: {
    media: {
      models: [
        /* shared list */
      ],
      image: {
        /* optional overrides */
      },
      audio: {
        /* optional overrides */
      },
      video: {
        /* optional overrides */
      },
    },
  },
}

#

Tutorial.step

ReferenceNodesMediaUnderstandingPage.steps.04.title

ReferenceNodesMediaUnderstandingPage.steps.04.p1

Json5
{
  type: "provider", // default if omitted
  provider: "openai",
  model: "gpt-5.2",
  prompt: "Describe the image in <= 500 chars.",
  maxChars: 500,
  maxBytes: 10485760,
  timeoutSeconds: 60,
  capabilities: ["image"], // optional, used for multi‑modal entries
  profile: "vision-profile",
  preferredProfile: "vision-fallback",
}
Json5
{
  type: "cli",
  command: "gemini",
  args: [
    "-m",
    "gemini-3-flash",
    "--allowed-tools",
    "read_file",
    "Read the media at {{MediaPath}} and describe it in <= {{MaxChars}} characters.",
  ],
  maxChars: 500,
  maxBytes: 52428800,
  timeoutSeconds: 120,
  capabilities: ["video", "image"],
}

CLI templates can also use:

- {{MediaDir}} (directory containing the media file)

- {{OutputDir}} (scratch dir created for this run)

- {{OutputBase}} (scratch file base path, no extension)

Tutorial.step

默认和限制

推荐默认：

ReferenceNodesMediaUnderstandingPage.steps.05.item1

ReferenceNodesMediaUnderstandingPage.steps.05.item2

- maxBytes：

- 画像：10MB

- 音声：20MB

- 動画：50MB

Rules:

ReferenceNodesMediaUnderstandingPage.steps.05.rule1

ReferenceNodesMediaUnderstandingPage.steps.05.rule2

ReferenceNodesMediaUnderstandingPage.steps.05.rule3

ReferenceNodesMediaUnderstandingPage.steps.05.rule4

#

Tutorial.step

ReferenceNodesMediaUnderstandingPage.steps.06.title

ReferenceNodesMediaUnderstandingPage.steps.06.p1

ReferenceNodesMediaUnderstandingPage.steps.06.p2

ReferenceNodesMediaUnderstandingPage.steps.06.p3

ReferenceNodesMediaUnderstandingPage.steps.06.item1

ReferenceNodesMediaUnderstandingPage.steps.06.sherpa

ReferenceNodesMediaUnderstandingPage.steps.06.whisperCli

ReferenceNodesMediaUnderstandingPage.steps.06.whisper

ReferenceNodesMediaUnderstandingPage.steps.06.item2

3. 提供商密钥

- 音声：OpenAI → Groq → Deepgram → Google

- 画像：OpenAI → Anthropic → Google → MiniMax

- 動画：Google

ReferenceNodesMediaUnderstandingPage.steps.06.disable

Json5
{
  tools: {
    media: {
      audio: {
        enabled: false,
      },
    },
  },
}

ReferenceNodesMediaUnderstandingPage.steps.06.note

Tutorial.step

機能（选项）

ReferenceNodesMediaUnderstandingPage.steps.07.p1

ReferenceNodesMediaUnderstandingPage.steps.07.p2

- openai、anthropic、minimax：画像

- google（Gemini API）：画像 + 音声 + 動画

- groq：音声

- deepgram：音声

ReferenceNodesMediaUnderstandingPage.steps.07.p3

ReferenceNodesMediaUnderstandingPage.steps.07.p4

Tutorial.step

ReferenceNodesMediaUnderstandingPage.steps.08.title

| 機能 | 提供商集成 | 注釈 |

| ---- |

ReferenceNodesMediaUnderstandingPage.steps.08.p3

| 音频 | OpenAI、Groq、Deepgram、Google | 提供商転写（Whisper/Deepgram/Gemini）。 |

| 视频 | Google（Gemini API） | 提供商视频理解。 |

Tutorial.step

推荐提供商

画像

ReferenceNodesMediaUnderstandingPage.steps.09.imageItem1

ReferenceNodesMediaUnderstandingPage.steps.09.imageItem2

音声

- ''openai/gpt-4o-mini-transcribe''、''groq/whisper-large-v3-turbo''、或 ''deepgram/nova-3''。

- CLI 回退：''whisper-cli''（whisper-cpp）或 ''whisper''。

- Deepgram 设置：''Deepgram（音声転写）''。

動画

ReferenceNodesMediaUnderstandingPage.steps.09.videoItem1

ReferenceNodesMediaUnderstandingPage.steps.09.videoItem2

Tutorial.step

附件策略

ReferenceNodesMediaUnderstandingPage.steps.10.p1

- ''mode''：''first''（默认）或 ''all''

ReferenceNodesMediaUnderstandingPage.steps.10.item2

- ''prefer''：''first''、''last''、''path''、''url''

ReferenceNodesMediaUnderstandingPage.steps.10.p2

Tutorial.step

设置例

#

Tutorial.step

1) 分钟享模型列表 + 覆盖

Json5
{
  tools: {
    media: {
      models: [
        { provider: "openai", model: "gpt-5.2", capabilities: ["image"] },
        {
          provider: "google",
          model: "gemini-3-flash-preview",
          capabilities: ["image", "audio", "video"],
        },
        {
          type: "cli",
          command: "gemini",
          args: [
            "-m",
            "gemini-3-flash",
            "--allowed-tools",
            "read_file",
            "Read the media at {{MediaPath}} and describe it in <= {{MaxChars}} characters.",
          ],
          capabilities: ["image", "video"],
        },
      ],
      audio: {
        attachments: { mode: "all", maxAttachments: 2 },
      },
      video: {
        maxChars: 500,
      },
    },
  },
}

#

Tutorial.step

ReferenceNodesMediaUnderstandingPage.steps.13.title

Json5
{
  tools: {
    media: {
      audio: {
        enabled: true,
        models: [
          { provider: "openai", model: "gpt-4o-mini-transcribe" },
          {
            type: "cli",
            command: "whisper",
            args: ["--model", "base", "{{MediaPath}}"],
          },
        ],
      },
      video: {
        enabled: true,
        maxChars: 500,
        models: [
          { provider: "google", model: "gemini-3-flash-preview" },
          {
            type: "cli",
            command: "gemini",
            args: [
              "-m",
              "gemini-3-flash",
              "--allowed-tools",
              "read_file",
              "Read the media at {{MediaPath}} and describe it in <= {{MaxChars}} characters.",
            ],
          },
        ],
      },
    },
  },
}

#

Tutorial.step

ReferenceNodesMediaUnderstandingPage.steps.14.title

Json5
{
  tools: {
    media: {
      image: {
        enabled: true,
        maxBytes: 10485760,
        maxChars: 500,
        models: [
          { provider: "openai", model: "gpt-5.2" },
          { provider: "anthropic", model: "claude-opus-4-5" },
          {
            type: "cli",
            command: "gemini",
            args: [
              "-m",
              "gemini-3-flash",
              "--allowed-tools",
              "read_file",
              "Read the media at {{MediaPath}} and describe it in <= {{MaxChars}} characters.",
            ],
          },
        ],
      },
    },
  },
}

#

Tutorial.step

ReferenceNodesMediaUnderstandingPage.steps.15.title

Json5
{
  tools: {
    media: {
      image: {
        models: [
          {
            provider: "google",
            model: "gemini-3-pro-preview",
            capabilities: ["image", "video", "audio"],
          },
        ],
      },
      audio: {
        models: [
          {
            provider: "google",
            model: "gemini-3-pro-preview",
            capabilities: ["image", "video", "audio"],
          },
        ],
      },
      video: {
        models: [
          {
            provider: "google",
            model: "gemini-3-pro-preview",
            capabilities: ["image", "video", "audio"],
          },
        ],
      },
    },
  },
}
Tutorial.step

状态输出

ReferenceNodesMediaUnderstandingPage.steps.16.p1

Terminal
📎 Media: image ok (openai/gpt-5.2) · audio skipped (maxBytes)

ReferenceNodesMediaUnderstandingPage.steps.16.p2

Tutorial.step

注意

ReferenceNodesMediaUnderstandingPage.steps.17.item1

ReferenceNodesMediaUnderstandingPage.steps.17.item2

ReferenceNodesMediaUnderstandingPage.steps.17.item3

Tutorial.step

関連文档

- ''设置''

ReferenceNodesMediaUnderstandingPage.steps.18.item2