What is codex-testing-kit?

Codex Testing Kit is a collection of tools built for OpenAI Codex to help software testers work faster, aimed specifically at the Vietnamese tester community. It packages ready made rules, reusable skills, detailed workflows, and prompt templates so Codex can assist with both manual testing and automation testing. The kit covers the testing process end to end: analyzing requirements, designing test cases, generating test data, building automation frameworks, writing automation scripts, testing APIs, dealing with flaky tests, and integrating with tools like Jira and Xray. It includes 26 specialized Codex skills that Codex can either choose automatically based on your request, or that you can call directly by typing a dollar sign followed by the skill name. The repository is organized into a few main folders. The .agents folder holds six sets of QA and automation rules plus the 26 skills. A plans folder contains step by step workflows for manual testing, automation testing, and cross module testing, each with its own quick start guide. A prompt_templates folder has ready to use prompt files for single tasks like generating requirements, creating a Playwright or Selenium framework, converting manual test cases into automation scripts, generating test data, and analyzing flaky tests. An AGENTS.md file at the root gives Codex repository level instructions on when to use which skill and rule. To use the kit, you either clone the repository or copy its .agents, AGENTS.md, plans, and prompt_templates folders into the root of the project you want to test. Then you open Codex, either the app or the command line version, point it at that project folder, and call a skill directly in your prompt, such as asking it to generate manual test cases from a requirements file. Codex reads AGENTS.md, loads the relevant skill file, and carries out the workflow. The README is written primarily in Vietnamese and does not state a license.