hailoc12/e2e_test_report_for_vibe_coding

This project is a testing skill that runs end to end tests against the live link of any AI generated, or vibe coded, product and automatically produces a polished quality assurance report. Rather than relying on a gut feeling that a product seems to work, it aims to give concrete test results, backed by a five layer test matrix, scores across seven criteria, a list of detected features, and screenshots as evidence for each conclusion. According to the README, the generated report is delivered as a DOCX or PDF file. Example output shown in the README includes a cover page with a large colored score badge, a scoring table across seven criteria with color coded rating cells and a highlighted total row, and detail cards explaining each individual criterion. If the tested product includes an AI assistant or a retrieval augmented chatbot, the skill also separately grades the accuracy of the AI responses by checking each answer against a source of truth, specifically to catch cases where the AI invents information that isn't true. The README's guiding idea is stated directly: don't trust what people say about a product, trust the test results. This positions the tool as a way for people who built something quickly using AI coding tools to get an honest, evidence backed assessment of whether their product actually works, before showing it to others or shipping it. The project is written in Python and is packaged as a skill, suggesting it is meant to be invoked by an AI coding assistant or agent framework rather than run as a standalone application. Because the excerpt does not include installation or usage commands, exact setup steps for running it are not shown here.