tpo-definition-of-done

Validates work items against Definition of Done checklists covering code, UX, testing, security, and release readiness.

Updated Mar 22, 2026
One-click install
npx skills add https://github.com/TECH-HY/SKILLS --skill tpo-definition-of-done-tech-hy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tpo-definition-of-done
Source: https://github.com/TECH-HY/SKILLS/tree/main/skills/tpo-definition-of-done
Command: npx skills add https://github.com/TECH-HY/SKILLS --skill tpo-definition-of-done-tech-hy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams and AI developer agents often mark work as complete when code is merely written, leading to untested features, missing error states, security gaps, and undocumented limitations reaching review or release. ## Core Features & Use Cases - Core Done Checklist: Enforces 15 mandatory checks covering acceptance criteria, scope control, tests, error handling, security, and documentation before any item is accepted. - Domain Completion Gates: Provides dedicated review tables for product, code quality, architecture, UX/UI states, testing evidence, security and data, documentation, and release readiness. - Structured Review Templates: Supplies completion report and final review templates so the AI Developer Agent reports evidence and the Technical Product Owner Agent records accept/reject decisions. - Use Case: After an AI agent implements a user story, run it through this Definition of Done to verify tests pass, UX states work, no secrets are logged, and produce a final Accepted/Rejected decision with required fixes. ## Quick Start Review the completed task against this Definition of Done and produce a final accept or reject decision with evidence.

Frequently Asked Questions about tpo-definition-of-done

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

FAQPage Schema
How do I write a Definition of Done for an agile team?

A Definition of Done lists mandatory quality criteria a work item must meet before acceptance, such as passing acceptance criteria, tests, code review, and documentation updates. This skill provides a ready checklist with Must items, evidence columns, and Not Done conditions.

How to review AI agent code output before accepting it?

Require the AI Developer Agent to submit a completion report listing files changed, test commands and results, UX evidence, deviations, and known limitations. Then verify each item against the Done checklist and record an Accepted, Rejected, or Accepted with follow-up decision.

What should a Definition of Done checklist include?

It should cover acceptance criteria, scope control, code quality, architecture conformance, UX/UI states, test evidence, security and data validation, documentation updates, and release readiness. Items that fail any Must check are explicitly Not Done.

Does this Definition of Done apply to bug fixes and refactoring?

Yes, it applies to user stories, technical tasks, bug fixes, UX/UI work, API changes, data model changes, refactoring, and release acceptance. The same Must checklist and evidence requirements govern all work types.

When is work considered not done even if code is written?

Work is Not Done when acceptance criteria fail, tests are missing without explanation, security risks remain unresolved, UX flows have dead ends, scope changed without approval, or the agent cannot explain what changed.