# Colossyan MCP > Colossyan MCP automates synthetic media production directly from your AI agent. Generate high-quality videos using saved templates, manage custom avatars, and track video renders without touching a traditional video editor. Get started by listing available actors or triggering a full JSON generation. ## Overview - **Category:** marketing-automation - **Price:** Free - **Tags:** ai-video, avatar-generation, synthetic-media, video-automation, text-to-video ## Description Generating professional videos used to involve complex workflows: uploading assets, manually adjusting variables, hitting 'render,' and then waiting hours for feedback. This MCP changes that. You connect Colossyan through Vinkius and handle the entire video lifecycle using simple natural language commands in your AI client. Need a new spokesperson? List available actors or create an instant avatar from an image link. Ready to make content? Trigger videos using existing templates, or submit full JSON descriptions for total creative control. The agent handles the technical rendering job, giving you real-time status updates and final public URLs when it's done. You can even manage your content library by retrieving details of finished videos or deleting old jobs you no longer need. ## Tools ### create_actor Builds an instant avatar using either an image or a video link. ### delete_job Deletes a specific video generation job from the platform. ### delete_video Removes a generated video asset entirely. ### generate_draft Creates an early draft version of a video using structured text data. ### generate_from_template Runs a video by populating variables within a pre-saved template. ### generate_video Generates a video manually using a full, detailed JSON description. ### get_job_status Checks the real-time status of any running video generation job. ### get_video Retrieves detailed information and metadata for a finished video asset. ### list_actors Pulls a list of all available avatars and actors in your workspace. ### list_voices Provides a catalog of all available voices across different languages. ## Prompt Examples **Prompt:** ``` List all available AI actors and voices in my Colossyan account. ``` **Response:** ``` I've retrieved your available assets. You have access to 15 Studio actors and 42 voices across 10 languages. Would you like to see the details for a specific language or actor type? ``` **Prompt:** ``` Generate a video from template 'tpl_9988' with the variable name set to 'Sarah'. ``` **Response:** ``` Starting video generation using template 'tpl_9988'. I've mapped the dynamic variable 'name' to 'Sarah'. The job ID is `vid_conf_123`. I'll monitor the status for you. ``` **Prompt:** ``` Check the status and get the link for video ID 'vid_conf_123'. ``` **Response:** ``` The video generation for `vid_conf_123` is finished! You can view it here: [Public URL]. The video is 45 seconds long and ready for download. ``` ## Capabilities ### Discover Virtual Talent List available actors, voices, and avatar types for your workspace. ### Create Custom Avatars Build instant avatars from existing images or video links. ### Automate Video Creation Generate full videos by using saved templates with dynamic variables or providing complete JSON descriptors. ### Manage Rendering Jobs Check the status of video jobs, retrieve finished content details, and delete assets when they are done. ### Generate Content Drafts Create preliminary video drafts directly from structured text summaries. ## Use Cases ### Personalized Onboarding Series An L&D specialist needs 50 versions of an onboarding video, each featuring a different employee's name. Instead of manually updating the template 50 times, they prompt their agent to use `generate_from_template` and pass a list of names as variables. ### Rapid Campaign Testing A marketing team needs to test how different voices sound for a campaign. They first call `list_voices` to see the options, then use those specific voice IDs when running `generate_video` with varied scripts. ### Client Asset Retrieval A producer finishes a video and needs to know its dimensions and public link. They prompt their agent to call `get_video` using the asset ID, getting all necessary metadata in one go. ## Benefits - Scale personalized campaigns: Use `generate_from_template` to inject dynamic variables into saved layouts, letting your AI client run hundreds of variations instantly. - Build talent on the fly: If you need a new spokesperson, use `create_actor` to turn an image or video link into a usable avatar right within the chat flow. - Maintain full control: When templates aren't enough, call `generate_video` and submit precise JSON descriptors for custom creative output. - Never get stuck waiting: Use `get_job_status` to monitor your renders in real time. You know exactly when your video is ready without refreshing a dashboard. - Clean up assets easily: When a project wraps, use `delete_video` or `delete_job` to keep your content library clean and organized. ## How It Works The bottom line is that your AI agent acts as the bridge between natural conversation and professional-grade media rendering. 1. Subscribe to this MCP and enter your Colossyan API Key. 2. Reference the tool in your AI agent (Claude, Cursor, etc.) by requesting a specific action, like listing available actors or generating content from a template. 3. Your agent sends the request to Colossyan; you get back status updates or final video links. ## Frequently Asked Questions **How do I list available avatars using the `list_actors` tool?** Just tell your agent to run `list_actors`. It pulls all current assets, letting you see every avatar and actor ID you can use for a video. **Does `generate_video` require the full JSON structure?** Yes. Using `generate_video` gives maximum control because it requires a detailed JSON descriptor, ensuring all necessary variables and settings are explicitly passed to Colossyan. **What is the difference between `delete_job` and `delete_video`?** `delete_job` cancels a rendering task that hasn't finished. `delete_video` removes a video asset that has already been successfully created. **Can I check status after running `generate_from_template`?** Absolutely. After triggering the render, use `get_job_status` with the provided job ID to monitor its progress until it's complete. **What kinds of input sources can I use when calling `create_actor`?** You can create an instant avatar using both image and video links. Just provide a URL for either format to generate the necessary assets. The tool handles the conversion from those inputs directly into your workspace avatars. **If my video generation fails, how do I use `get_job_status`?** The status response includes specific error codes if a job fails. You check the returned metadata for the reason and can then adjust your variables or input to retry the process. **What kind of structured data is needed when calling `generate_draft`?** You must provide a highly structured text summary with clear key-value pairs. The tool requires these specific inputs to map concepts accurately into a coherent video draft format. **After I use `get_video`, how do I ensure the final media is cleared using `delete_video`?** `delete_video` removes the finished rendered media file. This action targets only the asset, leaving the associated generation job record intact unless you also specifically call `delete_job`. **How can I check if my video has finished rendering?** You can use the `get_job_status` tool by providing the `videoId`. The agent will return the current status, such as 'finished', 'generating', or 'failed'. **Can I generate videos using my existing Colossyan templates?** Yes! Use the `generate_from_template` tool. You'll need the `templateJobId` and a JSON object for `dynamicVariables` to fill in the placeholders in your template. **How do I see which AI avatars are available in my account?** Simply ask the agent to run the `list_actors` tool. It will retrieve all available avatars, including Studio, Scenario, and Instant types configured for your workspace.