This turns Claude into a team lead that autonomously manages multiple AI agents until work is done. It runs a continuous loop: scans agent statuses, reads their progress, sends instructions, and coordinates dependencies without asking you to check in manually. The skill is opinionated about ownership (you drive the loop, never wait for permission to check again) and includes guardrails for what it can approve versus when to escalate. Useful when you've spun up several agents on related tasks and don't want to babysit the handoffs. The verify skill integration means it actually checks diffs and runs tests before calling something complete.
npx -y skills add codeaholicguy/ai-devkit --skill agent-orchestration --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
sickn33/antigravity-awesome-skills
moizibnyousaf/ai-agent-skills
github/awesome-copilot