Poster Design Studio

What Problem It Solves

Creating an eye-catching poster often involves a disconnect between the creative brief and the final visual output. Designers face repeated communication cycles to align on direction, or must handle technical tasks like asset compositing and style matching themselves. Poster Design Studio aims to bridge this gap by converting a user's initial idea—whether a text description, a product photo, or a draft—into a deliverable visual along the shortest path, reusing all decisions already established in the conversation.

Core Workflow

The skill is built around three clear execution routes, selected automatically or by user specification based on the input material:

1. Theme-to-Poster: Input is a text description only (e.g., event theme, promotional copy). The skill calls beatra.images.generate to transform the brief into a complete visual. This is the default starting point.
2. Photo-to-Poster: Input includes a source photograph. The skill uses beatra.images.transform to elevate the original image into a poster, preserving the subject while adding professional typography, style, and reserved text areas.
3. Draft Refinement: Input is an already-approved poster sketch. The skill employs beatra.images.edit for localized fixes or full-image adjustments without altering the overall layout.

Before execution, the skill guides the creation of a structured Poster Brief, with key elements including:
- Poster Type & Category (e.g., event poster, movie poster, promotional banner)
- Headline Information (as the visual anchor for composition)
- Focal Subject (performer, product, or key visual scene)
- Style Direction (e.g., "clean futuristic," "magazine bold")
- Canvas & Publishing Context (e.g., A3 print, 1:1 social media, 9:16 story)
- Reserved Text Area Preference (top, centered, or bottom)

Preparing the brief is free, but the remote calls for image generation are paid. The skill enforces a "Confirm and Execute Only Once" principle: before any call, it must display and lock in the final prompt, canvas, reference order, model, and all other parameters. Any parameter change constitutes a new paid job, requiring a new confirmation and client_request_id.

All remote operations must be performed via the bundled scripts/mcp_client.py, invoking specific tools (e.g., beatra.images.generate) with parameters passed as JSON via standard input. Using REST/OpenAPI as a fallback is prohibited. After generation, the skill polls the task using beatra.tasks.get and conducts a review based on visual impact, text readability, brand consistency, and canvas adaptation. It finally delivers the artifact link, actual dimensions, and billing credits charged.

Applicability & Constraints

• Technical Dependency: The skill strictly relies on its bundled MCP client and the Beatra platform's remote tools; other generation interfaces or models are not supported.
• Workflow Mode: Paid calls are atomic. Once confirmed and submitted, you cannot create a replacement job unless the original response status is genuinely unknown. If polling stalls or a cancellation returns 409, tracking continues on the original task without restarting it.
• Input Limitations: For the "Photo-to-Poster" route, uploading only transmits the file bytes; the skill does not automatically inspect the source photo's content, relying on the user's explicit annotation and subsequent visual review.
• Update Mechanism: The skill silently checks for updates (max once per 24 hours) and auto-installs newer versions. Updates come from fixed official addresses with file integrity checks; failure does not affect current usage or command execution.
• Best Practices: Users should establish theme, style, canvas, and other decisions early in the conversation to minimize interaction turns. For critical parameters like model selection (model=auto), the skill only queries beatra.models.list when it materially impacts the outcome.