> ## Documentation Index
> Fetch the complete documentation index at: https://docs.vidjutsu.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# VidJutsu — Video Intelligence API

> The feedback loop your agent is missing.

Watch, extract, transcribe, check, and **pre-flight platform compliance** on any AI-generated video or caption. Plus storage and tagging for your content library.

## Catch violations before you post

Every creator who's been demonetized, shadow-banned, or suspended knows the feeling. **VidJutsu Compliance** scans a video or caption against each platform's current Community Guidelines and ad policies, returns a deterministic risk score, and cites the exact policy clause behind every flag.

* **Four platforms** — YouTube, TikTok (consumer + ads), Instagram / Meta, Facebook Ads
* **Verbatim citations** — every violation links back to the platform's own policy page, with the exact quoted clause; you can verify against the live source
* **Deterministic scoring** — severity weights sum to a 0–100 score, capped at 100
* **Informational, not legal advice** — you still own the publish decision, we just surface what the platforms' own published text says

<Card title="Scan a caption in one line" icon="terminal" href="/cli/compliance">
  ```bash theme={null}
  vidjutsu compliance prompt \
    --text "Get rich in 30 days guaranteed — insider tip" \
    --platform instagram
  ```
</Card>

## Base URL

```
https://api.vidjutsu.ai
```

## Intelligence endpoints

Five endpoints that close the loop between generation and publishing:

<CardGroup cols={2}>
  <Card title="Watch" icon="eye" href="/api-reference/watch/watch">
    AI watches your video with a freeform prompt. Ask any question about what it sees.
  </Card>

  <Card title="Extract" icon="scissors" href="/api-reference/extract/create">
    Pull frames, audio, and metadata from any video via server-side processing.
  </Card>

  <Card title="Transcribe" icon="microphone" href="/api-reference/transcribe/create">
    Speech-to-text with word-level timing for every spoken word.
  </Card>

  <Card title="Check" icon="shield-check" href="/api-reference/check/check">
    Validate a spec against rules. QA gate before publishing.
  </Card>

  <Card title="Compliance" icon="scale-balanced" href="/cli/compliance">
    Pre-flight any video or caption against platform TOS. Cited policy clauses, deterministic risk score.
  </Card>
</CardGroup>

## Storage and CRUD

Organize your content library with accounts, posts, assets, references, and tags — all free.

<CardGroup cols={2}>
  <Card title="Accounts" icon="user" href="/api-reference/accounts/create">
    Pure data records for your social accounts. Tag and filter by niche, platform, or any custom key.
  </Card>

  <Card title="Posts" icon="file-video" href="/api-reference/posts/create">
    Track videos, captions, and briefs. Link posts to accounts and filter by tag.
  </Card>

  <Card title="Assets" icon="image" href="/api-reference/assets/create">
    Upload and manage media files — images, videos, and audio clips for your content pipeline.
  </Card>

  <Card title="References" icon="bookmark" href="/api-reference/references/create">
    Save reference URLs and inspiration links. Tag and retrieve by niche, format, or campaign.
  </Card>
</CardGroup>

## How it works

1. **Get your API key** — `POST /v1/subscribe` with your email, then pay the \$99/mo via Stripe. Poll `GET /v1/credits/status` after payment to retrieve your key.
2. **Analyze your content** — Call `/v1/watch`, `/v1/extract`, `/v1/transcribe`, `/v1/check`, or `/v1/compliance/*` on any public video URL or text input.
3. **Store and organize** — Create accounts, posts, assets, and references as data records. Tag them for filtering.

## Choose your interface

<CardGroup cols={2}>
  <Card title="REST API" icon="code" href="/api-reference/introduction">
    Direct HTTP calls to `api.vidjutsu.ai`. Works from any language or agent.
  </Card>

  <Card title="CLI" icon="terminal" href="/cli/index">
    Install the `vidjutsu` CLI and run intelligence calls from your terminal.
  </Card>
</CardGroup>