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

# Overview

> Open-source agents and managed workflow APIs for video search, video editing, and screenplay extraction.

Visual Agents bundles two open-source reference agents and a set of managed workflow APIs built on top of Visual Intelligence and Visual Search. Use the open-source code if you want to fork and customize; call the managed APIs if you want hosted endpoints that already work.

<CardGroup cols={3}>
  <Card title="Video Searching Agent" icon="magnifying-glass" href="/visual-agents/video-searching-agent">
    Open-source + managed API. Multi-platform discovery across YouTube, TikTok, Instagram, and X — query parsing, ranked references, SSE streaming.
  </Card>

  <Card title="Video Editing Agent (VEA)" icon="film" href="/visual-agents/video-editing-agent">
    Open-source + managed primitives. End-to-end pipeline: indexing → script generation → clip planning → TTS → render. Or call `/video/edit`, `/video/clip`, `/video/split` directly.
  </Card>

  <Card title="Screenplay Extraction" icon="clapperboard" href="/visual-agents/screenplay/submit-task">
    Managed API only. Submit a short drama episode → get back structured storyboard data (scene breakdowns, transcripts, character namelist). Async with webhook callback.
  </Card>
</CardGroup>

<Card title="Agent Examples" icon="book" href="/visual-agents/use-cases">
  Eight runnable ReAct agents — SOP compliance, service quality, security, public-platform search, video editing, personal memory, visual RAG, creator intelligence. Live-verified against `api.memories.ai`.
</Card>

## What's Inside

| Agent / Service           | Open-source repo                                                                                    | Managed API                                                            | Type                                         |
| ------------------------- | --------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------- | -------------------------------------------- |
| **Video Searching Agent** | [Memories-ai-labs/video-searching-agent](https://github.com/Memories-ai-labs/video-searching-agent) | `POST /queries/stream` ([API ref](/visual-agents/video-searching-api)) | Agent — searches social platforms            |
| **Video Editing Agent**   | [Memories-ai-labs/vea-open-source](https://github.com/Memories-ai-labs/vea-open-source)             | `POST /video/edit`, `POST /video/clip`, `POST /video/split`            | Agent + primitive editing endpoints          |
| **Screenplay Extraction** | — (managed service)                                                                                 | `POST /screenplay/tasks` (async) + status / cancel / download          | Workflow API — analyses short drama episodes |

> Looking for the **Memories CLI**? It's a separate ecosystem-wide tool — see the [Documentation overview](/home/overview#tools).

## Which One Do I Use?

| Goal                                                           | Pick                                                                                  |
| -------------------------------------------------------------- | ------------------------------------------------------------------------------------- |
| Discover trending content / creators across social platforms   | **Video Searching Agent** (or call `/queries/stream` directly)                        |
| Turn a long video into a short-form highlight reel             | **Video Editing Agent** (fork it for full pipeline)                                   |
| Just clip / split / re-edit individual segments                | Call the primitive endpoints under VEA (`/video/clip`, `/video/split`, `/video/edit`) |
| Extract structured storyboard / screenplay from drama episodes | **Screenplay Extraction**                                                             |
| Drive the platform from terminal / scripts                     | **Memories CLI**                                                                      |

## Two Ways to Adopt

| Path                           | When to use                                                      | What you do                                                            |
| ------------------------------ | ---------------------------------------------------------------- | ---------------------------------------------------------------------- |
| **Fork the open-source agent** | You want full control over prompts, tools, models, scoring logic | `git clone` the repo, configure keys, customize, run on your own infra |
| **Call the managed API**       | You just want results, no infra                                  | `POST` to the managed endpoint with your API key; we run the agent     |

The managed APIs and the open-source agents are kept in sync — the open-source code is the reference implementation of what the managed endpoint runs.

## Shared Foundation

* **Underlying APIs**: All agents/services call [Visual Intelligence](/visual-intelligence/getting-started/overview) (for inference) and optionally [Visual Search](/visual-search/introduction) (for indexed search).
* **Auth**: Same `Authorization: sk-mavi-...` API key as the rest of the platform — no separate key.
* **Billing**: Calls hit the underlying Visual Intelligence / Visual Search endpoints and bill against those line items. No additional "agent fee".
* **Webhooks**: Async services (Screenplay Extraction, VEA) require a configured webhook — see [Webhooks](/visual-intelligence/getting-started/webhooks).

## Getting Started

1. **Pick the agent or service** that matches your goal (table above).
2. **Get an API key** from the [Memories.ai Console](https://api-platform.memories.ai/stripe). The same key works across Visual Intelligence, Visual Search, and Visual Agents.
3. **Configure webhooks** if you'll use any async endpoint (VEA, Screenplay Extraction).
4. **Open the per-agent page** for setup instructions specific to that agent.