Wan 3.0 Video Prompt MCP
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Wan 3.0 Video Prompt MCP build a video prompt for a cat playing piano in a jazz club"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Wan 3.0 Video Prompt MCP
A deterministic, read-only Model Context Protocol server for building Wan 3.0 video prompts, planning short shot sequences, checking current WAN-30.video configuration boundaries, and returning canonical resources.
This community project supports the independent WAN-30.video workspace. It is not affiliated with or endorsed by Alibaba or the Wan project.
Tools
build_wan3_video_prompt— assemble a copy-ready production prompt.plan_wan3_shot_sequence— build a timed continuity plan for one to six shots.diagnose_wan3_video_prompt— check clarity plus current duration, resolution, reference, end-frame, and audio boundaries.get_wan3_video_resources— return canonical creation, guide, pricing, and policy URLs.
The server does not upload files, generate media, call model providers, access accounts, spend credits, or submit payments.
Related MCP server: chinavideoai
Run
npx -y github:gpt-img-2/wan3-video-prompt-mcpOptional environment variable:
WAN3_APP_BASE_URL=https://wan-30.videoOpenClaw
The installable Skill is in openclaw/wan-3-video-prompt-architect/SKILL.md.
openclaw mcp add wan3-video \
--command npx \
--arg -y \
--arg github:gpt-img-2/wan3-video-prompt-mcp \
--include 'build_wan3_video_prompt,plan_wan3_shot_sequence,diagnose_wan3_video_prompt,get_wan3_video_resources'If the installed OpenClaw release does not expose the mcp command group, configure npx -y github:gpt-img-2/wan3-video-prompt-mcp as a Stdio MCP server instead.
Development
pnpm install
pnpm validateRequires Node.js 20 or newer. Licensed under MIT.
Available Tools
4 toolsbuild_wan3_video_promptBuild a Wan 3.0 video promptARead-onlyIdempotent
Assemble a copy-ready text-to-video or image-to-video prompt from subject, action, scene, camera, lighting, audio, reference roles, and exclusions without calling a model.
Inputs: mode (required), action (required), subject (required), avoid, +7 more.
diagnose_wan3_video_promptDiagnose a Wan 3.0 video promptBRead-onlyIdempotent
Check prompt clarity and WAN-30.video's currently displayed duration, resolution, reference-count, end-frame, and audio boundaries without submitting generation.
Inputs: mode (required), prompt (required), resolution (required), durationSeconds (required), +3 more.
get_wan3_video_resourcesGet WAN-30.video resourcesBRead-onlyIdempotent
Return canonical WAN-30.video creation, guide, pricing, and policy resources without fetching external content.
Inputs: topic.
plan_wan3_shot_sequencePlan a timed Wan 3.0 shot sequenceARead-onlyIdempotent
Turn one to six short shots into a timeline and continuity checklist, reporting when shot durations do not match the requested clip length.
Inputs: shots (required), totalDurationSeconds (required), continuityAnchors.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 4 tools
Each tool serves a clearly distinct phase: fetching resources, building a prompt, planning shot sequences, and diagnosing prompt constraints. Although build and diagnose both relate to prompts, their actions and outputs are clearly separated by the descriptions.
All tool names follow a consistent snake_case verb_noun pattern with a shared wan3_video namespace. The convention is predictable across all four tools.
Four tools is well-scoped for a focused prompt-assistance server. Each tool covers a meaningful part of the prompt workflow without unnecessary overlap or bloat.
The surface covers resource lookup, prompt assembly, shot planning, and prompt diagnosis, which are the core needs for a WAN 3.0 prompt tool. Minor gaps exist around iterative refinement, saving, or exporting generated prompts, but the core workflow is supported.
Maintenance
Related MCP Connectors
Transcode, host and caption video from a prompt. Fifteen tools, nine read-only, nothing deletes.
Plan, compare, price, generate, and recover AI video from compatible MCP clients.
Free, open source library of AI video prompts that really worked, each next to the video it produced and credited to its creator. Describe a video in plain words, in any language: your AI finds the prompts that fit by meaning, shows you the videos and can adapt the best one to your idea. Read only, with no account and no key.
Hosted video MCP server for branded explainers and reference-driven edits.
Related MCP Servers
- AlicenseAqualityDmaintenanceEnables deterministic, read-only design of OpenSora video prompts and shot plans, including prompt diagnosis and access to workflow guides, without generating videos or requiring API keys.4MIT
- AlicenseAqualityDmaintenanceEnables deterministic, read-only creation of bilingual video prompts, timed reference-aware shot plans, and focused prompt diagnostics without API keys or external provider calls.4MIT
- AlicenseAqualityDmaintenanceProvides read-only tools to build English or Chinese video prompts, plan reference-aware sequences, diagnose prompts, and fetch canonical SeeVido resources without API keys.4MIT
- AlicenseAqualityDmaintenanceEnables deterministic, read-only construction of Seedance-oriented video prompts, planning of multimodal reference-aware sequences, prompt diagnosis, and discovery of canonical See Dance 2 resources.4MIT