Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@io.github.Headdao/slidemaster-mcp Create a video about Mars colonization"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP (Model Context Protocol) Server for the SlideMaster Public API. This server exposes 20 tools that let any MCP-compatible AI assistant create AI-powered presentation videos from a simple topic.
Sign in to SlideMaster and generate a key at Settings > API Keys.
Add the following to your claude_desktop_config.json:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"slidemaster": {
"command": "npx",
"args": ["-y", "@slidemaster/mcp-server"],
"env": {
"SLIDEMASTER_API_KEY": "your-api-key-here"
}
}
}
}After saving the config, restart Claude Desktop. You should see the SlideMaster tools available in the tool picker.
Related MCP server: mcp-ToseaAI
# Install dependencies
npm install
# Build TypeScript
npm run build
# Run the server (stdio transport)
SLIDEMASTER_API_KEY=your-key npm startVariable | Required | Description |
| Yes | Your SlideMaster API key |
| No | Override API base URL (default: |
Tool | Method | Endpoint | Description |
| POST |
| Generate a presentation outline from a topic |
| POST |
| Create a new project |
| POST |
| Initialize a file upload session (PPTX/PDF) |
| POST |
| Mark an upload as complete |
| POST |
| Render AI-generated slide images |
Tool | Method | Endpoint | Description |
| POST |
| Generate a narration script for one slide |
| POST |
| Generate scripts for all slides in a project |
| POST |
| Generate text-to-speech audio |
| POST |
| Compile the final video |
Tool | Method | Endpoint | Description |
| GET |
| List all projects (paginated) |
| GET |
| Get project details |
| PATCH |
| Update project properties |
| DELETE |
| Delete a project |
| GET |
| List slides in a project |
| PATCH |
| Update a slide's title or script |
| DELETE |
| Delete a slide |
Tool | Method | Endpoint | Description |
| GET |
| Check project processing status |
| GET |
| Export project with download URLs |
| GET |
| List available TTS voices |
Tool | Method | Endpoint | Description |
| POST |
| End-to-end: topic to finished video |
User: Create a 5-slide presentation about renewable energy in Taiwan
Claude: I'll create that presentation for you using SlideMaster.
[Calls generate_outline with topic="renewable energy in Taiwan", slides_count=5]
Here's the outline I generated:
1. Taiwan's Energy Landscape
2. Solar Power Expansion
3. Offshore Wind Development
4. Government Policy & Targets
5. Future Outlook
Let me now create the project and render the slides.
[Calls create_project with title="Renewable Energy in Taiwan"]
[Calls render_slides with the outline data]
[Calls batch_generate_scripts]
[Calls generate_tts]
[Calls generate_video]
Your video is being generated! Let me check the status.
[Calls check_status]
The video is ready! Here are your download links:
[Calls export_project]
- Video: https://...
- Slides: https://...Alternatively, use the one-shot pipeline:
User: Make me a video about machine learning basics
Claude: I'll use the topic-to-video pipeline to handle everything in one step.
[Calls topic_to_video with topic="machine learning basics"]
[Polls check_status until complete]
[Calls export_project for download links]
Your video is ready!Full API documentation is available at https://slidemaster.tw/api-docs.
MIT
Generate narration scripts for ALL slides in a project in one batch operation.
Inputs: project_id (required), slide_ids, regenerate, prompt_style.
Check the processing status of a project. Use this to poll for completion after triggering render, TTS, or video generation.
Inputs: project_id (required).
Create a new SlideMaster project. Returns the project_id needed for subsequent operations.
Inputs: title (required), type, description.
Permanently delete a project and all associated slides, audio, and video files.
Inputs: project_id (required).
Delete an individual slide from a project.
Inputs: slide_id (required).
Export project data and download URLs for all generated assets (slides, audio, video).
Inputs: project_id (required).
Generate a presentation outline from a topic. Returns a list of slide titles and descriptions that can be used to create slides.
Inputs: topic (required), style, language, slides_count.
Generate a narration script for a single slide using AI.
Inputs: slide_id (required), regenerate, prompt_style.
Generate text-to-speech audio. Provide project_id (all slides) or slide_id (single slide).
Inputs: slide_id, project_id, voice_name, tts_provider, +1 more.
Compile the final video from rendered slides and TTS audio. This is the last step in the pipeline.
Inputs: project_id (required), fps, force, resolution.
Get detailed information about a specific project, including its metadata and current status.
Inputs: project_id (required).
List all projects in the account.
Inputs: limit.
List all slides belonging to a specific project, including their titles, scripts, and render status.
Inputs: project_id (required).
List all available TTS voices across all providers. Useful for choosing a voice before generating TTS.
Inputs: no named parameters.
Render AI-generated slide images for a project. Each slide gets a background image produced by the image generation model.
Inputs: slides (required), project_id (required), image_size.
One-click pipeline: create a presentation from a topic. Default produces slides only. Set include_scripts=true for narration scripts, include_video=true for full MP4 video. Poll check_status afterward.
Inputs: topic (required), style, language, tts_voice, +5 more.
Update project properties such as title, description, TTS provider, voice, language, or speaking rate.
Inputs: project_id (required), title, tts_voice, description, +3 more.
Update an individual slide's title or narration script.
Inputs: slide_id (required), title, script.
Mark a file upload as complete so the server can begin processing the uploaded presentation.
Inputs: project_id (required).
Initialize a file upload session for an existing presentation file (PPTX/PDF). Returns a pre-signed upload URL and project_id.
Inputs: title (required), filename (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 20 tools
Tools are mostly mapped to distinct resources and actions: project CRUD, slide updates, script/TTS/video generation, and status polling. The main ambiguities are get_project vs check_status (both expose status) and the generate_outline vs topic_to_video entry points when starting from a topic.
The naming convention is largely consistent snake_case verb_noun across the set (list_projects, update_slide, generate_tts, delete_project). Minor deviations like topic_to_video and batch_generate_scripts break the strict pattern but remain predictable.
20 tools is at the high end for a single server and feels heavy, even for the multi-stage presentation-to-video workflow. Several granular operations (upload_init/upload_complete, generate_script/batch_generate_scripts) could have been combined into parameterized tools.
The main topic-to-video pipeline is covered end to end, and project lifecycle tools are complete. However, slide management lacks a create/add-slide operation, and generate_outline has no explicit follow-up tool to turn the outline into slides, leaving a notable gap for step-by-step workflows.
MCP server for Clipkit — gives AI agents a video toolbox via the Clipkit schema.
Model Context Protocol server for the Apideck Unified API. Connect any MCP-compatible agent framework to 100+ accounting systems, HRIS platforms, file storage providers, and more through one integration. More information https://www.apideck.com/mcp-server
MCP server that lets AI assistants use all OneSchema features exposed via the public API.
MCP server for progressive tool usage at any scale (see https://klavis.ai)
