Use Defuddle when an agent needs to turn a URL or saved HTML page into cleaned article content, Markdown, or JSON metadata as part of a repeatable research, clipping, summarization, indexing, or archival workflow. Invoke it instead of using a browser or web clipper manually when the operator needs deterministic command-line extraction, structured metadata, or batch-friendly page cleanup. The scope boundary is Defuddle’s parser and CLI: the skill is about extracting readable main content and metadata from pages, not listing a generic browser extension, SDK, or content platform.