# CaptionHub MCP

> CaptionHub manages your entire video localization pipeline. Use this MCP to track projects, trigger AI auto-transcription for videos, and approve subtitle sets directly from any connected agent. It handles the full caption lifecycle—from initial project creation to final export.

## Overview
- **Category:** image-video
- **Price:** Free
- **Tags:** subtitles, transcription, video-localization, ai-transcription, subtitle-management, workflow-automation

## Description

Need to manage subtitles across multiple languages? This MCP connects CaptionHub to your agent so you can handle video localization workflows without leaving your toolchain. You'll start by listing all projects and checking their metadata, figuring out exactly what needs work. From there, you can trigger AI-powered transcription directly against a video asset in the workspace. Once the transcript is ready, you use natural conversation to approve or archive specific caption sets for different languages. If you need final files, retrieving export links is simple. Everything runs through your agent, making sure every project status and webhook is monitored automatically. This kind of deep control over complex content workflows is what Vinkius makes possible, letting you manage the entire subtitle process conversationally.

## Tools

### approve_captions
Marks a specific set of captions as approved for use.

### archive_project
Removes an entire video localization project permanently from the system.

### create_project
Sets up a brand new captioning project for incoming video content.

### export_captions
Generates and retrieves the final download URL for finished captions in various formats.

### get_account_info
Retrieves basic information about your connected CaptionHub account.

### get_project_details
Pulls specific details and status reports for one video project by its ID.

### list_projects
Returns a list of every active video captioning project you manage.

### list_webhooks
Shows all the webhook endpoints currently set up to monitor your projects.

### transcribe_video
Starts an AI process to automatically transcribe a video asset for a specific project.

### update_project
Changes metadata, like the title or target language, on an existing project.

## Prompt Examples

**Prompt:** 
```
List all my current video projects in CaptionHub.
```

**Response:** 
```
I've retrieved your projects. You have 5 active projects, including 'Product Keynote' (ID: proj_1) and 'Training Module v2' (ID: proj_2).
```

**Prompt:** 
```
Start auto-transcription for the 'Product Keynote' project.
```

**Response:** 
```
Auto-transcription has been triggered for 'Product Keynote' (ID: proj_1). I will monitor the progress and notify you when it's ready for review.
```

**Prompt:** 
```
Get the French subtitle export link for project proj_123.
```

**Response:** 
```
Retrieving French caption export... Your download link is ready: https://api.captionhub.com/exports/abcd-1234.srt
```

## Capabilities

### Check Project Status
List all video captioning projects or retrieve specific metadata for one project ID.

### Generate Transcriptions
Trigger AI auto-transcription on a selected video asset within your workspace.

### Manage Subtitles
Approve, archive, or modify caption sets for specific languages using simple commands.

### Get Export Files
Retrieve the direct download link needed to export finished captions in various formats.

### Set Up Monitoring
Monitor active webhooks and project statuses to ensure your localization pipeline runs smoothly.

## Use Cases

### The handover review
A video editor finishes a draft in Spanish (ID: proj_123). Instead of emailing the Localization Manager to check status, they ask their agent to `get_project_details` for proj_123. The manager gets the status and confirms it's ready to be reviewed.

### Mass cleanup
A project is canceled because the source video was pulled. Instead of manually finding and deleting records, the content producer runs `archive_project` on the relevant ID to clean up the system immediately.

### Localization audit
The manager needs a list of all projects that are currently set up for German translation. They run `list_projects`, which returns detailed metadata, allowing them to quickly see exactly what’s active and where the assets are.

### Final delivery prep
The team just finished all reviews on a project. The content producer asks their agent for the export link using `export_captions`, receiving the direct URL needed to hand off the final files.

## Benefits

- You instantly check project status. Instead of jumping between tabs to see if a job is done, you just ask your agent to list all projects or get details for one using `get_project_details`.
- Transcription happens on demand. Triggering auto-transcription via the `transcribe_video` tool means you don't wait until a batch process runs; you start it exactly when you need it.
- Approvals are conversational. Use natural language to approve or archive captions. No more clicking through multiple approval screens just to mark something 'Done'.
- Exporting is simple. Need the SRT file? The `export_captions` tool gets you a direct download link immediately, bypassing complex download menus.
- It keeps your pipeline visible. By monitoring webhooks and project statuses, you know if an automated step failed or succeeded without manual checks.

## How It Works

The bottom line is that your agent treats CaptionHub like another built-in function, letting you manage complex video assets through conversation alone.

1. First, connect your API key to the MCP. This gives your agent access to all CaptionHub functions.
2. Next, use natural conversation—like 'List my current video projects'—to pull up project statuses and details.
3. Finally, when you need transcripts or export links, simply ask your agent to perform the action for a specific ID.

## Frequently Asked Questions

**How do I check all my video projects using list_projects?**
You simply ask your agent to use `list_projects`. It returns a full manifest of every active project ID and its current status, so you know exactly what's in the system.

**Can I approve captions using approve_captions?**
Yes. You direct your agent to use `approve_captions` along with the specific language code. This officially marks the set as approved for use by everyone else on the team.

**What is the best way to get a new project started? Use create_project.**
To start fresh, you tell your agent to `create_project`. This initializes all necessary metadata and gives you a clean slate for the video content.

**Does transcribe_video only work on finished videos?**
No. You trigger `transcribe_video` directly from your workspace against the asset ID. It runs AI auto-transcription to get you a usable transcript for review immediately.

**If I need detailed metadata for a single video, what does `get_project_details` provide?**
It gives you comprehensive data on one specific project. You can pull details like source/target languages or assigned owners without listing all projects first.

**How do I manage webhooks or monitor status changes using `list_webhooks`?**
You use `list_webhooks` to see every active webhook. This lets your agent confirm which endpoints are monitoring project statuses, ensuring reliable workflow automation.

**What format can I expect when I call `export_captions`?**
You receive a specific download URL for the finished captions. The link is ready to pull the data in various standard formats, like SRT or VTT files.

**If I need to modify the details of an existing project, how does `update_project` work?**
You use `update_project` to change metadata for a running job. This function lets you adjust things like ownership or target language without recreating the entire workflow.

**Can I trigger auto-transcription for a video using the agent?**
Yes! Use the `transcribe_video` tool with the Project ID. Your agent will trigger CaptionHub's AI engine to start transcribing the video content instantly.

**How do I get the download link for finished subtitles?**
Simply ask the agent to `export_captions` and provide the Project ID and the language code (e.g., 'fr' or 'es'). It will return the export URL for your finished caption set.

**Does the integration allow approving captions via the workspace?**
Yes. Use the `approve_captions` tool with the Project ID and Language. This will mark the specific caption set as approved in your CaptionHub workflow.