DLazy Tongyi Wanxiang 2.7 Video Generation
Paste the following prompt into your AI chat to install this skill:
Please follow https://skillhub.cn/install/skillhub.md to install @user_f95f21b8/dlazy-wan2-7 into your AI assistant.
About this skill
Problem
When agents call video models, the friction is often operational rather than model-related: authentication, uploading local media, polling async jobs, and handling balance or permission errors. wan2.7 wraps Tongyi Wanxiang 2.7 video generation as a dlazy wan2.7 call, covering text-to-video, first/last-frame-to-video, and reference-to-video tasks.
How It Works And Boundaries
The skill is a thin client over dLazy's hosted API. On execution, prompts and parameters are sent to api.dlazy.com; local image, video, or audio paths are uploaded to files.dlazy.com so the model can read them; generated output is usually a URL hosted on files.dlazy.com. Authentication uses a dLazy API key, configured through the CLI or supplied via DLAZY_API_KEY, and the key can be rotated or revoked in the dashboard.
- Synchronous calls wait for the result;
- Async calls with
--no-waitreturntask.generateIdand status, requiring status polling; - Error handling must distinguish
unauthorized,insufficient_balance, missing parameters, missing local files, and safety-policy rejections.
It fits engineers who already manage dLazy organization keys and credits, not fully local inference, private deployment, or scenarios that must avoid uploading media.
Use Cases
- Generate short product clips from copy inside an agent workflow using Tongyi Wanxiang 2.7.
- Upload local first and last frame images to the dLazy API to produce a fixed-start preview video.
- Create showcase shots for an ecommerce product page by driving video generation from a reference image.
- Poll an async generateId returned by --no-wait and check balance or permission errors.
Best For
- Backend engineers calling video APIs in CI or agents, who want unified auth, media upload, and async status handling.
- Creative product engineers turning reference images into showcase videos, who want to validate the generation flow via CLI.
- Platform engineers managing dLazy organization keys and credits, who need to distinguish balance, permission, and safety-policy errors.
- Application engineers building multimodal agents, who want to upload local first/last frames and poll generation results.
Related Skills
A ComfyUI image-generation skill that uses a five-step dialog to collect prompts, prefill templates, confirm parameters, submit API jobs, and return results.
A design-system guidance skill for Impeccable that produces token-based rules, component states, accessibility criteria, and QA checklists.
Director-level video lapian that diagnoses material precision, then produces frame evidence, director analysis, a style bible, and a showcase video.
Plan and brief Amazon MAIN, Listing, and A+ image sets from verified product facts, then return plan and image QA status.