skills.rest
microsoft
microsoftOFFICIAL
Explore all skills by this author →
Website𝕏@OpenAtMicrosoftGitHub

hve-builder

Author, review, and validate GitHub Copilot prompt-engineering artifacts through staged evidence-backed workflows.

1.4k|284|Updated Nov 2, 2025
Search Tags:#Validation#Skills#Prompt Engineering#Workflow#AGENTS#artifact review#gitHub copilot
One-click install
npx skills add https://github.com/microsoft/hve-core --skill hve-builder
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: hve-builder
Source: https://github.com/microsoft/hve-core/tree/main/.github/skills/hve-core/hve-builder
Command: npx skills add https://github.com/microsoft/hve-core --skill hve-builder

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Creating and maintaining high-quality Copilot prompts, instruction files, agents, subagents, and skills is error-prone: artifacts drift from best practices, reviews are subjective, and behavior changes ship untested. This Skill provides one evidence-backed lifecycle that creates, improves, refactors, replaces, reviews, or validates prompt-engineering artifacts with static review rubrics, behavior testing, and host validation gates.

Core Features & Use Cases

  • Six lifecycle modes: create, improve, refactor, replace, review, and validate, each with stage gates, narrow write boundaries, and explicit Pass/Revise/Deferred/Blocked outcomes.
  • Evidence-grounded quality standard: applies a ranked requirements catalog distilled from OpenAI, Anthropic, and Google prompting guidance, plus a bounded static-review rubric with severity-graded findings.
  • Behavior testing and research delegation: routes Major mutations through the hve-builder-tester skill and all open-ended codebase exploration through the rpi-research bridge.
  • Use Case: A team wants to add a new Copilot subagent for Terraform module review. The Skill selects the right artifact type, authors it against the requirements catalog, runs fresh-context static review, executes behavior tests, and returns a verdict with evidence logs.

Quick Start

Ask the agent to use the hve-builder skill to review the prompt file at .github/prompts/deploy.prompt.md against the requirements catalog and report severity-graded findings.

Frequently Asked Questions about hve-builder

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a new Copilot agent or subagent with hve-builder?▼

Invoke the skill in create mode with the target path and requirements. It selects the artifact type by responsibility and load timing, authors against the requirements catalog, runs static review, and executes behavior testing for major mutations before reporting a verdict.

How do I review a Copilot prompt file for quality issues?▼

Use review mode, which is read-only for source artifacts. A fresh-context static review applies the bounded rubric dimensions, assigns Critical/High/Medium/Low severities to findings, and returns a Pass, Revise, or Blocked verdict with the smallest resolving changes.

What artifact types does hve-builder support?▼

It handles prompts, instruction files, agents, subagents, and skills. Type selection is based on responsibility, activation surface, load timing, and authority rather than a fixed preference ladder, and existing artifacts are reused before new ones are created.

Can a host project extend hve-builder with its own conventions?▼

Yes. Instruction files extend it via applyTo globs, skills via semantic description matching, and subagents via routing descriptions and host registration. Extensions add scoped conventions but cannot redirect the workflow, widen write scope, or weaken safety rules.

What happens when behavior testing cannot run?▼

A behavior verdict of Not available resolves the run to Deferred with an exact rerun condition. Eligible Minor or Medium mutations and no-runtime review targets may instead record a satisfied-and-skipped behavior gate with an explicit reason.

Related Skills

Explore More Software Engineering Skills→
awieczorkawieczork

artifact-creator

Generate and refactor VS Code Copilot customization artifacts as structured markdown files.

Community
Advanced
pavel-molyanovpavel-molyanov
244

skill-test-designer

Design test scenarios for AI skills with acceptance criteria and model selections.

Community
Intermediate
anilcancakiranilcancakir
3

claude-md-rules-creator

Generate and refactor CLAUDE.md and path-scoped rule files.

Community
Advanced
MrJmpl3MrJmpl3

0065-laravel-policies-and-authorization

Centralize Laravel access control using Policies and Gates.

Community
Basic