five-steps-agents

What problem does it solve?

Convert informal requirements for coordinated workflows into a robust, deployable multi-agent architecture that enforces explicit agent boundaries, measurable signal tiers, comprehensive error handling, isolated toolsets, and cost-aware model routing to avoid god-agents and silent failures.

Core Features & Use Cases

• Structured Intake & Decomposition: Guided discovery questions and automatic decomposition into distinct responsibility-based agents with an identified orchestrator.
• Five-Step Agent Architecture: For each agent, apply Boundaries, Signal Tiers, Error Handling, Tool Handling, and Model Routing to produce ready-to-deploy orchestrator and sub-agent artifacts.
• Generation, Validation, and Deployment: Produce orchestrator.md and sub-agent .md files, run a 12-point validation checklist (minimum 10/12), and emit deployment artifacts such as route commands and a Cortex entry.
• Use Cases: Orchestrating lead-finding pipelines, client onboarding flows, content production queues, and competitive intelligence systems where 2+ distinct responsibilities must coordinate reliably.

Quick Start

Provide a one-sentence use case description and target deployment context so the skill can decompose responsibilities and generate orchestrator plus sub-agent prompts.