DeepSeek Harness Plugin

WaveSpeedAI/wavespeed-dsh-skill

Stars ★ 3 Category Skills Added 2026-08-20

Generate and edit AI media (image, video, audio, 3D) with WaveSpeed models via the open-source wavespeed CLI: live catalog search, per-model input-schema introspection, local-file upload with @path markers, and price checks before running.

Install

# from GitHub (first run asks for allowBuilds approval — follow the hint, retry)

dsh plugin --profile web add github:WaveSpeedAI/wavespeed-dsh-skill

Any plugin you install runs third-party code with your own permissions — it can read your files, use your credentials, and reach the network, and tool approvals don’t sandbox it. GitHub-sourced plugins also run build scripts at install time — pnpm blocks those until you allow them, so an install can stop with ERR_PNPM_GIT_DEP_PREPARE_NOT_ALLOWED or ERR_PNPM_IGNORED_BUILDS; dsh prints the exact key to add under allowBuilds in your profile’s pnpm-workspace.yaml, and the install works on the next run. Allowing a build is a trust decision: only install sources you trust, and pin a commit (github:owner/repo#sha).

README

中文

A DeepSeek Harness (dsh) skill that lets the agent generate and edit AI media — image, video, audio, 3D — through the WaveSpeed platform, using the open-source @wavespeed/cli.

For images and videos the skill uses the CLI's shortcuts (wavespeed generate|edit|extend, defaults Seedream 5 Pro and Seedance 2.5), and the WaveSpeed MCP tools when they are connected. Every other model on the platform is one wavespeed run <model-id> call: the skill teaches the agent to browse the live catalog, read any model's input schema, and execute it — including uploading local files with the @path marker.

Requirements

  • Node.js ≥ 18
  • The WaveSpeed CLI: npm install -g @wavespeed/cli
  • A WaveSpeed account — wavespeed login opens the key page at wavespeed.ai and handles the rest

Install

As a dsh plugin (recommended — the bundle registers the skill automatically):

dsh plugin --profile web add github:WaveSpeedAI/wavespeed-dsh-skill

Or copy the skill directory by hand — project-level (checked into the repo, shared with your team):

mkdir -p .dsh/skills
git clone --depth 1 https://github.com/WaveSpeedAI/wavespeed-dsh-skill /tmp/wss \
  && cp -r /tmp/wss/skills/wavespeed .dsh/skills/ && rm -rf /tmp/wss

User-level (available in every project):

mkdir -p ~/.dsh/skills
git clone --depth 1 https://github.com/WaveSpeedAI/wavespeed-dsh-skill /tmp/wss \
  && cp -r /tmp/wss/skills/wavespeed ~/.dsh/skills/ && rm -rf /tmp/wss

Other agents (Claude Code, Cursor, Codex): the CLI installs the same skill directly —

wavespeed skill install

Packaged plugins with the same skill and MCP server: Claude Code, Codex, Gemini CLI, Kimi Code CLI, OpenCode.

What the agent can do with it

# text → image (default model: Seedream 5 Pro)
wavespeed generate image "a cyberpunk skyline at golden hour" --json

# edit a local image (uploaded automatically via @path)
wavespeed edit image "replace the background with a sunlit kitchen" --image @./input.jpg --json

# image → video (default model: Seedance 2.5); price it first, nothing is submitted
wavespeed generate video "subtle parallax" --first-frame @./hero.jpg --estimate --json

# any other model on the platform
wavespeed run <model-id> -p "..." --json

The skill also covers project aliases (wavespeed.json), schema introspection, price/balance checks, and prediction history.

License

MIT — same as the CLI.


WaveSpeed AI — AI image & video generation platform. Try it in the browser: Image generator · Video generator

Content from the project README on GitHub ↗

Links

More in this category

View the whole category →

Community comments

Comments are public GitHub Discussions. Loading them connects to GitHub and Giscus; a GitHub account is required to post.