bria-fibo-image

What it does

Instructs an agent to build and operate rights-aware Bria FIBO and FIBO Lite image-generation workflows, using structured prompts, a reference image, asynchronous status handling, webhooks, and cost gates, with safe artifact downloads. Applies when a user requests Bria/FIBO generation, refinement, inspiration, reproducibility, hosted API integration, or a licensed-data image workflow; excludes FIBO Edit, video, generic image editing, or third-party FIBO gateways.

How it works

• Uses Bria's first-party hosted v2 API unless the user asks for local weights; keeps local-weight licensing separate from hosted API terms. Never route a first-party request through fal.ai, Replicate, or another gateway.
• Establishes decision paths and routes via v2 OpenAPI endpoints, selecting among image/generate, image/generate/lite, structured_prompt/generate, and structured_prompt/generate/lite, with designated fields and behaviors for each route. Enforces the exact route schema (one of the allowed combinations) per request.
• Documents and applies an evidence vocabulary with defined labels (Documented fact, Provider claim, Operational heuristic, Unknown).
• Requires explicit start with a decision record detailing intent, delivery, inputs, outputs, governance, and spend before writing code or submitting work.
• Specifies exact routes, required fields, and behavior differences for high-fidelity vs faster iteration vs structured prompts; handles prompts, images, seeds, and structured prompts with attention to licensing and data handling.
• Provides a complete hosted example client script and price guidance, including per-call costs and rate limits, while enforcing client-side caps and validation.

When to use it

• Use when a user requests Bria/FIBO image-generation, refinement, inspiration, reproducibility, hosted API integration, or licensed-data image workflows. Do not use for FIBO Edit, video, generic image editing, or third-party FIBO gateways.

What it can touch

• Interacts with hosted Bria API at base URL https://engine.prod.bria-api.com/v2 via routes: /image/generate, /image/generate/lite, /structured_prompt/generate, /structured_prompt/generate/lite. Includes fields such as resolution, steps_num, negative_prompt, prompt, images, structured_prompt, webhook_url, output_type, and other standard image-generation inputs. The skill notes that the OpenAPI accepts image input as a public URL or raw base64 without data:image prefix and supports JPEG/PNG/WEBP; it enforces validation on field combinations and rejects unknown fields locally.

Caveats

• Volatile facts and sources were checked 2026-07-10; recheck schemas, pricing, plan limits, model cards, and governing terms before production.
• Privacy consequences indicate certain data paths and subprocessor details are not fully established; users should confirm contracts/DPAs before handling sensitive inputs. Lite uses hosted data transfer to Bria; local/on-prem requires a separately licensed deployment.
• The current schemas enforce asynchronous operation by default and may incur hosted-usage costs; approvals and costs are governed by Bria’s pricing and contract terms.
• The skill emphasizes that structured-prompt routes do not expose image outputs for some endpoints and that unknown fields should be rejected locally.