Telegram media and attachments

### Photo albums

Send multiple image attachments in one `message` tool call. OpenClaw groups consecutive photos into Telegram albums of up to 10 images, in their original order. Automatic replies with multiple photos use the same grouping. A single photo, including a final remainder of one, is sent separately.

{
  action: "send",
  channel: "telegram",
  to: "123456789",
  message: "Trip photos",
  attachments: [
    { type: "image", media: "https://example.com/photo-1.jpg" },
    { type: "image", media: "https://example.com/photo-2.jpg" },
  ],
}

The caption goes on the first photo. Text that exceeds the caption limit follows the album as a separate message. Photos sent as documents, other media types, and messages with inline buttons keep separate sends so their existing controls and delivery behavior are preserved.

### Audio messages

Telegram distinguishes voice notes from audio files. Default: audio-file behavior; tag `[[audio_as_voice]]` in the agent reply to force a voice-note send. Inbound voice-note transcripts are framed as machine-generated, untrusted text in agent context, but mention detection still uses the raw transcript so mention-gated voice messages keep working.

{
  action: "send",
  channel: "telegram",
  to: "123456789",
  media: "https://example.com/voice.ogg",
  asVoice: true,
}

### Video messages

Telegram distinguishes video files from video notes. Video notes do not support captions; provided message text sends separately.

{
  action: "send",
  channel: "telegram",
  to: "123456789",
  media: "https://example.com/video.mp4",
  asVideoNote: true,
}

### Locations and venues

Use the existing `send` action with one standalone `location` object. Coordinates send a native pin; adding both `name` and `address` sends a native venue card. Location sends cannot be combined with message text or media.

{
  action: "send",
  channel: "telegram",
  to: "123456789",
  location: {
    latitude: 48.858844,
    longitude: 2.294351,
    accuracy: 12,
    name: "Eiffel Tower",
    address: "Champ de Mars, Paris",
  },
}

### Stickers

Inbound: static WEBP is downloaded and processed (placeholder `<media:sticker>`). When sticker media is unavailable, including animated TGS and video WEBM, OpenClaw uses the sticker's emoji as the message text. If no emoji is available, it uses `<media:sticker>`. Cached descriptions take precedence over this fallback.

Sticker context fields: `Sticker.emoji`, `Sticker.setName`, `Sticker.fileId`, `Sticker.fileUniqueId`, `Sticker.cachedDescription`. Descriptions are cached in OpenClaw SQLite plugin state to reduce repeated vision calls.

Sticker descriptions use the configured `agents.defaults.imageModel` before shared automatic image-model selection, including the provider's MiniMax image routing. The sticker description uses one selected model and does not try the configured fallback list if that model fails. A failed description is not cached; general media analysis can still run separately with its normal fallback handling.

Enable sticker actions:

{
  channels: {
    telegram: {
      actions: {
        sticker: true,
      },
    },
  },
}

Send:

{
  action: "sticker",
  channel: "telegram",
  to: "123456789",
  fileId: "CAACAgIAAxkBAAI...",
}

Search cached stickers:

{
  action: "sticker-search",
  channel: "telegram",
  query: "cat waving",
  limit: 5,
}