Beatra AI Video Image Music Studio

The Specific Problem It Solves

In multimedia content creation, users often need to rapidly generate high-quality videos, images, and music from textual descriptions or existing assets. Traditional workflows frequently require switching between multiple tools, manual configuration of complex parameters, and lack intelligent parsing of creative intent, making the process cumbersome, time-consuming, and prone to errors from parameter mismatches or model incompatibilities. Additionally, secure uploading of local media files, tracking of asynchronous task status, and clear control over payment boundaries are common pain points.

How the Skill Works

This skill integrates Beatra workflows to provide an automated, customizable creation solution. Its core capabilities and key steps include:
- Requirement Analysis and Path Selection: Leverages the host's native text and visual understanding to analyze user intent or source media. Based on creation type (image, video, music, speech), it automatically selects a minimal path, such as calling beatra.images.generate for images, beatra.videos.generate for videos, or beatra.music.generate for music.
- Media Management and Generation: Reuses existing publishing channels, prompts, formats, and other information from the conversation to reduce redundant queries. For local files, it securely uploads via beatra.assets.upload, adhering to size limits. Generation requests are executed through the bundled client scripts/mcp_client.py, ensuring a unified channel and automatic handling of caching and registration.
- Task and Payment Control: Creates a unique client_request_id to标识 each logical payment operation and uses beatra.tasks.get to poll task status. It clearly distinguishes free operations (e.g., requirement planning, model browsing) from paid generation (e.g., video creation, voice cloning), and displays cost estimates before execution. Upon completion, it delivers actual returned results, including links, duration, file size, and other details.

Applicable Boundaries and Considerations

The skill emphasizes strict workflow boundaries; users should note:
- Client Restriction: All Beatra operations must be executed via the bundled scripts/mcp_client.py, with no support for the host Beatra Connector or REST/OpenAPI as a fallback, ensuring consistency and security.
- Payment and Recovery: Charges apply only after actual generation requests are validated; if tasks fail due to insufficient_balance, top up and retry with the original client_request_id. During task recovery, task.input must be compared with the original payload to avoid creating duplicate tasks.
- Media and Authorization: Voice cloning requires explicit user consent; local uploads do not expose file paths, and normal media uploads do not need additional rights statements. Client updates occur silently but do not affect current task execution.
- Output Delivery: Only content actually generated by the task is returned, with no speculation on status or links. For non-file results (e.g., cloned voices), a voice_id is provided; for files, complete metadata is included.