dbx-skill-architect

Name: dbx-skill-architect
Availability: InStock
Author: DBvc

Package recurring user requests into testable agent skills with triggers and eval artifacts.

5|1|Updated Jan 18, 2026

Search Tags:#governance #skill-architecture #agent-skills #evals #patch-hypothesis #domain-substance

Ready to run

npx skills add https://github.com/DBvc/skills --skill dbx-skill-architect

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: dbx-skill-architect
Source: https://github.com/DBvc/skills/tree/main/skills/dbx-skill-architect
Command: npx skills add https://github.com/DBvc/skills --skill dbx-skill-architect

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill helps teams transform recurring user requests into bounded, testable agent skill packages with clear triggers, outputs, and evaluation artifacts, reducing ambiguity and risk in automation design.

Core Features & Use Cases

Boundary-first design: defines mode/route/operation and hard gates before building any artifact.
Domain substance and evaluation: supports domain-discovery gates, domain content contracts, and runner/eval artifacts.
Patch-first improvement: promotes patch hypotheses over full rebuilds when improving existing skills.

Quick Start

Draft a new skill skeleton with a short, concrete domain scenario and open a patch plan if improvements are needed.

Frequently Asked Questions about dbx-skill-architect

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

FAQPage Schema

How do I design reusable agent skills with bounded scope and testable outputs?▼

Design reusable agent skills by defining explicit triggers, outputs, and evaluation artifacts before building. Boundary-first design establishes mode, route, and operation hard gates to ensure repeatable, auditable skill deployment.

What is patch-first improvement for existing agent skills?▼

Patch-first improvement promotes patch hypotheses over full rebuilds when refining existing agent skills. This approach applies targeted modifications and domain-discovery gates to update skill packages without discarding the original architecture.

How do I evaluate agent skills when domain specifics are incomplete or uncertain?▼

Evaluate agent skills with incomplete domain specifics by applying domain-discovery gates and domain content contracts. These mechanisms generate runner and eval artifacts that validate skill behavior under uncertain conditions.

Does boundary-first design enforce safety checks and artifact validation for agent skills?▼

Boundary-first design enforces hard gates, safety checks, and artifact validation for agent skills. This ensures repeatable, auditable skill deployment by validating outputs before any artifact is finalized.

Can I draft a new skill skeleton from a concrete domain scenario and open a patch plan?▼

You can draft a new skill skeleton from a short, concrete domain scenario and open a patch plan for improvements. This generates a bounded, testable agent skill package with clear triggers and outputs.

What are the limitations of patch-first improvement for agent skill architecture?▼

Patch-first improvement for agent skill architecture requires existing skill packages to patch. When domain substance is entirely missing, a full boundary-first design rebuild is necessary to establish triggers, outputs, and eval artifacts.