skills.rest
DeepJH
DeepJHCOMMUNITY
Explore all skills by this author →
GitHub

doubao-product-qa

Converts PRDs, prototypes, and code into traceable QA baselines, test cases, bugs, and release decisions.

2|Updated Aug 9, 2026
Search Tags:#test automation#traceability#api testing#test cases#qa testing#bug analysis#release decision
One-click install
npx skills add https://github.com/DeepJH/doubao-skill-and-info --skill doubao-product-qa-deepjh
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: doubao-product-qa
Source: https://github.com/DeepJH/doubao-skill-and-info/tree/main/skills/doubao-product-qa
Command: npx skills add https://github.com/DeepJH/doubao-skill-and-info --skill doubao-product-qa-deepjh

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve? It turns scattered product requirements, prototypes, interfaces, code, and test records into a single auditable QA evidence chain, so test plans, execution results, bugs, and release judgments stay consistent and traceable instead of being assembled ad hoc. ## Core Features & Use Cases - Controlled QA Workflow: A stage controller (bootstrap, anchor, complete) manages baseline, design, execution, change, and release phases with gate checks that block incorrect counts, verdicts, or P0 classifications. - Multi-Platform Testing Support: Covers Web, API, iOS, Android, and WeChat mini-program testing with environment detection scripts, device matrix runners, OpenAPI-driven test manifest generation, and manual handoff packages when automation is blocked. - Evidence-Based Bug and Release Discipline: Enforces S1-S4 severity and P0-P3 priority rules, preserves first-failure evidence, and only allows go/no-go verdicts backed by formal execution evidence. - Use Case: Given a PRD and a web app URL, the skill builds a requirement-risk-case traceability model, executes P0 journeys, files bugs with reproduction evidence, and delivers a release decision report as a document or spreadsheet. ## Quick Start Ask the AI to create a test plan and execute QA for your feature by providing the PRD or project directory and the target URL or build.

Frequently Asked Questions about doubao-product-qa

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

FAQPage Schema
How do I create a test plan from a PRD automatically?▼

Provide the PRD file or URL and the skill bootstraps a QA run, decomposes requirements into risk mechanisms, and generates traceable test cases with priorities. The controller then validates coverage counts before delivering the plan as a document or spreadsheet.

How do I generate API test cases from an OpenAPI spec?▼

Run the generate_api_manifest.py script with your OpenAPI or Swagger file to produce an operation manifest and coverage matrix. It creates happy-path, missing-parameter, and unauthorized test cases per operation, with write methods gated behind explicit environment authorization.

Can I run automated tests on iOS from a Windows or Linux machine?▼

No, XCUITest and iOS Simulator require macOS. On Windows or Linux hosts the skill detects the limitation and generates a manual test guide and handoff package with step-by-step cases instead of pretending automation ran.

What testing frameworks does it support for API automation?▼

It supports pytest with httpx for Python projects, Supertest with Vitest for Node.js projects, and REST Assured with JUnit 5 for Java projects. Framework selection follows detection of the existing project stack rather than forcing a new one.

Why does the QA gate block my release decision?▼

The gate blocks verdicts that lack decisive execution evidence, have mismatched P0 and blocking-bug sets, or show inconsistent coverage counts. Fix the listed BLOCK items in qa-run.json and rerun the same complete command to proceed.

When should a failure not be filed as a formal bug?▼

Precheck runs, single small samples, and isolated observations stay as pending confirmation until evidence reaches the required grade. Environment, data, and automation-script failures are classified separately from product defects.

Related Skills

Explore More Software Engineering Skills→
pisithrpspisithrps

code-first-draft

Generate initial feature implementations from PRDs and existing codebases.

Community
Advanced
zillabunnyzillabunny

fix-issues

Coordinate batch bug-fixing workflows across trackers, GitHub issues, and agent worktrees.

Community
Advanced
tclemtclem
11

investigate-alert

Diagnose production alerts by aggregating logs, metrics, traces, and source code.

Community
Advanced
Logos-LiberLogos-Liber
17

test-frameworks

Select and implement testing frameworks for web, mobile, API, and unit tests.

Official
Basic