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

# For your AI

> Hand your videngineer report to Claude, ChatGPT, or another assistant and turn it into a script or production brief.

For your AI answers: **how do I give a teardown to Claude, ChatGPT, or another assistant without losing the structure?**

This is the agent-facing page. The rest of the docs avoid model names; this page names them because its purpose is handoff.

There are two ways to do it. Copy and paste, documented below, works everywhere and needs no setup. Or skip the pasting entirely: [the connector](/agents/mcp-connector) gives your assistant read access to your own teardowns over MCP, so it fetches the hook, the beats, or one single cut itself. See [Connector recipes](/agents/mcp-worked-examples) for what that looks like in practice.

## The manual flow

<Steps>
  <Step title="Analyze a video">
    Run a teardown in the app, or open a public teardown. See [Your first teardown](/quickstart).
  </Step>

  <Step title="Collect the structural pieces">
    Grab the hook, emotional arc, pacing, beat sheet, CTAs, shot mix, and any Cut Blueprints you generated.
  </Step>

  <Step title="Paste them with your context">
    Add your product, audience, desired CTA, constraints, and production limits. Ask the assistant to apply the pattern to your story.
  </Step>
</Steps>

```text theme={null}
I analyzed a reference video and want to write my own version for [MY PRODUCT].

From the teardown:
- Hook: [hook text] - type [hook type], strength [1-10]
- Re-hook cadence: [timing]
- Emotional arc: [arc]
- Beat sheet: [beat labels with timestamps]
- Shot mix: [Blueprint summary]
- Sound direction: [Sound layer summary]

My product: [2 sentences]
My audience: [who they are, their pain, their goal]
CTA: [what they should do]
Constraints: [length, budget, format, production limits]

Write a 60-second script that uses the same structure with my story.
Format: VO lines with [VISUAL NOTE] cues for each scene.
```

<Note>
  Apply the pattern, never the content. The teardown gives you the structure that works; your story makes it yours.
</Note>

## Useful report pages for assistants

<CardGroup cols={2}>
  <Card title="Script analysis" icon="chart-line" href="/layers/script-analysis">
    Hook, re-hooks, beat sheet, CTAs, phrases, and voice-over.
  </Card>

  <Card title="The Blueprint" icon="map" href="/layers/blueprint">
    Shot mix, pacing, camera language, palette, and production scope.
  </Card>

  <Card title="The Cuts" icon="scissors" href="/layers/cuts">
    Playable clips, timing, and Cut Blueprint behavior.
  </Card>

  <Card title="Scene prompts" icon="image" href="/scriptwriters/recreation-prompts">
    Where to use scene prompts and the Cut Blueprint API endpoint.
  </Card>
</CardGroup>

## Feeding these docs to your assistant

Every page in these docs is assistant-ready. Use the contextual menu on any page to **copy the page as markdown** or **view it as markdown**, then paste the relevant context into Claude, ChatGPT, or your preferred assistant.

If your assistant needs field definitions for a layer, paste the relevant layer page into its context:

* [Blueprint](/layers/blueprint)
* [The Cuts](/layers/cuts)
* [Script analysis](/layers/script-analysis)
* [Sound layer](/layers/sound-layer)
* [Cast & World](/layers/cast)

## API note

For API work, use the API key from Account. Most agent work does not need the raw API: [the connector](/agents/mcp-connector) wraps the same access in twenty-one tools, eighteen of which are free reads. Cut Blueprint is available at `POST /api/v1/project/{job_id}/cut-brief` with `{"technique_id": "cut-001"}`; the endpoint returns `202 running` while generating and the stored blueprint once ready.