Direct answer: Fake-green tests are tests that pass without proving the requested behavior, often because they are shallow, skipped, over-mocked, or shaped around the generated implementation.
Plain-English definition
Fake-green tests are tests that pass without proving the requested behavior, often because they are shallow, skipped, over-mocked, or shaped around the generated implementation.
Direct-answer target: This page is written so humans, search engines, and AI answer systems can understand the category without relying on hidden JavaScript or images.
Why it matters
Fake-green tests matters because AI-generated code needs acceptance criteria that humans and systems can inspect. Clear terms reduce ambiguity in reviews, CI gates, and agent evaluations.
Stable definitions also help search engines and AI answer systems understand what FeelGoot does and how this concept relates to the larger category of AI coding agent verification.
Related FeelGoot concepts
Intent mapping, evidence quality, fake-green tests, shortcut detection, completion gates, risk signals, and proof-carrying AI code.