GitHub Spec Kit

AI agent briefs,
development-ready.

Converts plain-language AI agent briefs into complete development packages.

A repeatable, auditable path from AI concept to development-ready specification. No technical translation required.

View the Specification ↓

From brief to build-ready

📋

Specify

Define requirements with user stories, acceptance criteria, and functional requirements — structured for engineering handoff.

📐

Plan

Generate implementation plans with technical architecture and phasing strategy — giving teams a clear roadmap before a single line is written.

Tasks

Break specifications into actionable work items with acceptance criteria — so every task has a clear definition of done.

Chief Estimator Risk Agent

USD 285M Project Regional Acute Care Hospital Phoenix, Arizona

To demonstrate this workflow, a Chief Estimator Risk Agent brief was run through the full GitHub Spec Kit process. The result is a complete specification for an agent that queries SharePoint historical estimate data and generates executive-ready conceptual risk briefings.

The output package includes the agent file, prompt file, output contract, and validation rubric — along with a live example risk assessment for a USD 285M Regional Acute Care Hospital Expansion project in Phoenix, Arizona. The assessment surfaces 9 estimate risks with historical evidence, likelihood and impact ratings, and mitigation strategies.

Everything the team needs

The complete specification package — from requirements through implementation tasks.

Core Specification

Complete requirements, user stories, and acceptance criteria for the Risk Estimation Agent system.

View Specification →

Example Risk Assessment

Live demonstration for a USD 285M Regional Acute Care Hospital Expansion — executive-ready output with 9 risks, evidence, and mitigations.

View Example Output →

Research & Findings

Background research, technology decisions, and design rationale behind the Risk Estimation Agent.

View Research →

Implementation Plan

Technical architecture, phasing strategy, and development approach for building the agent.

View Plan →

Data Model

Entity definitions, relationships, and data structures powering the Risk Estimation Agent system.

View Data Model →

Quick Start Guide

How to get started with the Risk Estimation Agent — step-by-step setup instructions.

View Quick Start →

Implementation Tasks

Detailed breakdown of all work items and deliverables with acceptance criteria for each task.

View Tasks →

Enterprise AI Governance

This methodology reflects the approach applied in a large-scale enterprise AI program to assess and govern 70 AI use cases. It gives non-engineering leaders a repeatable, auditable path from AI concept to development-ready specification — without requiring technical translation at every step.

GitHub Spec Kit is free and open source. Fork it, adapt it, and use it to bring structure to your own AI development programs.