This skill wraps Qwen Coder CLI delegation into Claude workflows when you explicitly ask for it. It handles the plumbing: checking auth, formatting English prompts, selecting models (Qwen2.5-Coder or QwQ), setting approval modes from plan-only to yolo, and parsing results back into your session. The setup is careful about permissions and won't auto-apply code changes without confirmation unless you specifically ask for yolo mode. Use it when you want a second opinion on architecture, need QwQ's deep reasoning for security analysis, or want to continue a previous Qwen session without leaving Claude. The structured output format and metadata tracking make it easy to compare approaches between the two systems.
npx -y skills add giuseppe-trisciuoglio/developer-kit --skill qwen-coder --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
juliusbrussee/caveman
mattpocock/skills
shadcn/improve
obra/superpowers
forrestchang/andrej-karpathy-skills
vercel-labs/skills