wendy — Claude Code skill: what it does, how to install
Guides agents through building and deploying software to WendyOS edge hardware like Raspberry Pi or Jetson. It handles container permissions and Swift remote debugging.
npx skills add wendylabsinc/claude-skills --skill wendy -a claude-codeWhat wendy does
The skill instructs coding assistants on working with WendyOS, an embedded Linux operating system designed for edge hardware such as Raspberry Pi 4/5 and NVIDIA Jetson. It details the wendy CLI, creates project scaffolds with wendy init, and deploys binaries using wendy run. Structured output via --json is consistently applied to avoid interactive terminal prompts.
Swift is the default programming language for the platform, utilizing automated ARM64 cross-compilation. Building applications in any other language requires a custom Dockerfile. Device access rules, including host network access, cameras, audio, and Jetson GPU acceleration, are defined in wendy.json.
Interactive troubleshooting runs through a GDB server on port 4242 using wendy run --debug, which attaches to CodeLLDB inside VS Code. Application telemetry points to an on-device OpenTelemetry collector on port 4318 via HTTP protobuf. Hardware communication requires the wendy CLI preinstalled on the host system, which connects over gRPC port 50051.
What it helps with
- deploy Swift applications to Raspberry Pi or NVIDIA Jetson boards
- configure hardware and network permissions in
wendy.json - set up remote GDB debugging on port 4242
- route logs and traces to the local OpenTelemetry collector
Example prompts for the agent
- Initialize a new Wendy Lite project for Swift and enable camera entitlements in wendy.json
- How do I set up remote debugging on my Jetson board using VS Code?
- Connect the edge board to WiFi using wendy device wifi connect
How to install wendy
1Skills CLI (Vercel)
npx skills add wendylabsinc/claude-skills --skill wendy -a claude-code2Manually: put the files in place
~/.claude/skills/wendy/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/wendy/
1Skills CLI (Vercel)
npx skills add wendylabsinc/claude-skills --skill wendy -a codex2Manually: put the files in place
.agents/skills/wendy/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add wendylabsinc/claude-skills --skill wendy -a cursor2Manually: put the files in place
.cursor/skills/wendy/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install wendylabsinc/claude-skills wendy2Manually: put the files in place
.github/skills/wendy/SKILL.md1Skills CLI (Vercel)
npx skills add wendylabsinc/claude-skills --skill wendy -a gemini-cli2Platform CLI
gemini skills install https://github.com/wendylabsinc/claude-skills --path wendyManually: put the files in place
.devin/skills/wendy/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add wendylabsinc/claude-skills --skill wendy -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about wendy
How to install wendy in Claude Code?
Run npx skills add wendylabsinc/claude-skills --skill wendy -a claude-code. You can also place the instruction file directly into ~/.claude/skills/wendy/SKILL.md.
What does the wendy skill do?
It gives the AI complete reference instructions for WendyOS and the wendy CLI. It helps manage edge boards, compile Swift for ARM64, and configure container entitlements.
Can you use languages other than Swift?
Yes, but non-Swift code requires a dedicated Dockerfile. Built-in automated cross-compilation in the CLI is configured specifically for Swift Package Manager projects.
