dLazy Veo 3.1 Video Generation

Problem This Skill Solves

This skill is for engineers who need to turn text, images, or mixed media inputs into usable video clips. The core problem is not local rendering; it is how to call Google Veo 3.1 through a stable interface while handling cloud credentials, file uploads, async tasks, and structured errors. It wraps the dLazy hosted API as a CLI workflow, making it easier to reuse in automation scripts, agent toolchains, or content pipelines.

How It Works

The skill is a thin client over the dLazy hosted API. When invoked, prompts and generation parameters are sent to the dLazy API for inference. If inputs include local image, video, or audio paths, the CLI uploads them to dLazy media storage so the model can read them. Generated results are typically returned as hosted URLs. It supports synchronous waiting and --no-wait async mode, where the response includes a task ID and status for later polling. API keys can be scoped to a dLazy organization, supplied via environment variables, and rotated or revoked from the dashboard. Common failures include missing credentials, missing required parameters, local file read errors, insufficient balance, and async task failures.

Boundaries and Cautions

The skill depends on an external SaaS service, so it is not suitable for offline generation, strict data-residency constraints, or private model deployments. Local files pass through dLazy media storage, and output URLs are hosted remotely. Before production use, review service terms, source materials, safety policies, and credit management.