Plain-English definition
Intent drift is the mismatch between a user’s original software task and the code change an AI agent actually produced.
Why it matters
Intent drift 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.