Break any complex task into a plan, spawn sub-tasks, track progress, and loop until done — multi-step agent orchestration without the chaos.
Gives Claude a structured "plan → execute → verify → loop" workflow for long-running or multi-step tasks: decomposes the goal into a numbered task list, executes sub-tasks one at a time with status tracking, validates each result before advancing, handles partial failures with retries or replan, and produces a final summary with all artefacts. Heavily inspired by the Superpowers/agentic-dev patterns that dominate skills.sh trending charts. Works best when paired with MCP tools for file I/O, web search, or code execution.
No account needed · Instant delivery
Hand-picked Claude skills + tips for getting more out of your AI tools. No spam, unsubscribe anytime.
Weekly tips on using Claude effectively. Unsubscribe anytime.