Rpi

This is the full-cycle orchestrator that runs discovery, crank, and validation in strict sequence without letting you skip phases or compress the workflow. It enforces TDD by default, reruns validation up to three times on failure, and won't let acceptance tests slide. The phase isolation model is real: each step runs through bounded skill invocations, not inline shortcuts. Complexity detection scales the pre-mortem and vibe checks from inline reviews to full council mode. If you want autonomous end-to-end execution with guardrails instead of ad-hoc task running, this is the entry point. The tradeoff is deliberate: slower but actually completes the loop.

skills.sh installs: 1.5K
GitHub stars: 454
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add boshu2/agentops --skill rpi --agent claude-code

Installs into .claude/skills of the current project.

Files

RPI

Own the authorized outcome through finish. RPI (research, plan, implement) is one native pass, not fixed phases: use the native coding agent and shell. BD or the caller's tracker owns work and handoffs; Git owns content and delivery. AgentOps supplies a small charter and one fresh judgment where a mistake is costly, not a scheduler.