Classify request intent, risk, and uncertainty before approving project mutations.
npx skills add https://github.com/rpelizza/PelizzAI --skill pelizzai-router
Please help me install this Agent Skill. Skill: pelizzai-router Source: https://github.com/rpelizza/PelizzAI/tree/main/.agents/skills/pelizzai-router Command: npx skills add https://github.com/rpelizza/PelizzAI --skill pelizzai-router
This skill solves the problem of unmanaged agent behavior by acting as a central orchestrator that classifies task risk, intent, and uncertainty before any action is taken. It prevents unauthorized mutations and ensures every project change is ratified by the user.
Use the pelizzai-router skill to classify my request and propose a safe execution route for this project.
High-intent search queries and answers about installing and using this skill.
Agent task orchestration prevents unauthorized code modifications by acting as a central router that classifies task risk, intent, and uncertainty. It enforces a mandatory kickoff gate so the user must approve the proposed execution route before any persistent state mutation occurs.
A safety gate in agent workflow governance is a mandatory checkpoint that requires user ratification before proceeding with mutating tasks. It ensures the user reviews and approves the proposed route and artifacts, preventing the agent from executing unauthorized project modifications.
To route an agent request for a greenfield project, the orchestrator automatically identifies high uncertainty and mandates a discovery and planning phase. This ensures a safe execution route is proposed and ratified by the user before the agent can write any code.
Yes, you can use this agent router for bug resolution and feature implementation workflows. The orchestration mechanism applies to all development workflows by classifying request intent, risk, and uncertainty to determine the safest execution path for any project modification.
The primary limitation of automated agent task classification is its strict requirement for a kickoff gate, proposal analysis, and user ratification. The agent cannot bypass these safety gates or mutate persistent state without explicit user approval, which may slow down rapid prototyping.
hujianbestDefine project scope and risk during HarnessFlow's Frame stage.
opensage-agentBuild throwaway terminal applications to test state machines and UI variations.
radenadriMap a codebase through phase-based analysis of structure, dependencies, and data flows.
rpelizzaClassify and route coding tasks through a mandatory verification lifecycle.
