Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.veedcrawl.com/llms.txt

Use this file to discover all available pages before exploring further.

All five tools are registered automatically the moment the MCP server starts — no additional configuration required. Your agent can call any of them by name once the server is connected to your host.
Fetch structured metadata for any public video URL. Returns synchronously with no polling.
​
url
string
required
Public video URL from YouTube, TikTok, Instagram, X/Twitter, or Facebook.
Returns: platform, title, description, author info, view/like/comment/share counts, duration, thumbnails, tags, and publish timestamp.
Return the complete transcript for a video. The server handles all async polling internally — your agent receives the finished result when the job is done.
​
url
string
required
Public video URL or direct media file URL.
​
mode
string
default:"auto"
Transcription strategy:
  • native — use platform-provided captions only
  • generate — run AI speech-to-text only
  • auto — try native captions first, fall back to AI speech-to-text
​
lang
string
Language hint for transcription, e.g. en, es, ur. Optional.
Returns: completed transcript job object with full text and timestamps.
Ask any question about a video and receive a structured answer. Optionally constrain the output shape with a JSON Schema. The server handles all async polling internally.
​
url
string
required
Public video URL or direct media file URL.
​
prompt
string
required
What to extract or analyze from the video. Up to 4096 characters.
​
lang
string
Language hint for transcription, e.g. en, es, ur. Optional.
​
schema
object
JSON Schema to constrain the structure of the extraction output. Optional.
Returns: completed extraction job object with the structured result.Example prompts:
  • "Extract the hook, main argument, and call to action"
  • "List every product mentioned with the timestamp it appears"
  • "What claims does the speaker make? Rate each one as factual, opinion, or unverified"
  • "Rewrite this as a Twitter thread"
Fetch a TikTok creator’s public profile snapshot. Returns author info, total video count, and recent videos with full per-video engagement stats.
​
username
string
TikTok username, with or without the leading @. Required if url is not provided.
​
url
string
Full TikTok profile URL, e.g. https://www.tiktok.com/@creator. Required if username is not provided.
​
limit
number
default:"12"
Number of recent videos to return. Accepts 1–24.
Returns: author bio and stats, plus an array of recent videos. Each video object includes views, likes, comments, shares, caption, hashtags, thumbnail, and publish timestamp. You can pass any video’s URL directly into get_video_transcript or extract_from_video.
Fetch an Instagram creator’s public profile snapshot. Returns verified status, follower and following counts, total post count, and recent content.
​
username
string
Instagram username, with or without the leading @. Required if url is not provided.
​
url
string
Full Instagram profile URL, e.g. https://www.instagram.com/creator/. Required if username is not provided.
​
limit
number
default:"12"
Number of recent posts to return. Accepts 1–24.
Returns: author bio, verified flag, follower/following/post counts, and an array of recent posts. Each post includes its type (video, post for images, or sidecar for carousels), along with views, likes, comments, caption, and publish timestamp. Video posts also include a direct videoUrl you can pass into get_video_transcript.

​
Chaining tools together

The profile tools are designed to feed directly into the video tools. Here is a complete example that audits a creator without you supplying a single video URL:
1

Fetch the profile

Call get_instagram_profile with a username to get the creator’s recent posts and engagement stats.
get_instagram_profile(username="hubermanlab", limit=24)
→ 283M followers, verified, last 24 posts with view/like counts
2

Get the transcript

Pass a videoUrl from the profile response into get_video_transcript.
get_video_transcript(url=top_video.videoUrl)
→ Full transcript with timestamps
3

Extract structured data

Pass the same URL into extract_from_video with a targeted prompt.
extract_from_video(
  url=top_video.videoUrl,
  prompt="List every health claim made and the evidence cited"
)
→ Structured list of claims with citations