Obsidian Web Clipper

Obsidian Web Clipper templates are structured JSON files that define how the clipper extension captures web content and metadata. Every variable climbs the extraction ladder: schema, then meta, then selector, then prompt, in decreasing order of reliability.

Context needed

Ask the user about anything missing below until you understand what they want to capture. Never assume.

1. Vault convention: How the user organizes their vault, folder structure, naming pattern, and any vault-wide rule the template needs to follow. read: CLAUDE.md, AGENTS.md, README.md if present
2. Existing frontmatter systems: What properties (YAML frontmatter) does the user already use across notes? Any property they always include? Any they want for this specific content type? read: .obsidian/templates.json for the templates location, .obsidian/plugins/templater-obsidian if the user uses Templater and configures their template location there, .obsidian/types.json for property types the user has already defined
3. Note body structure: What should the captured content look like? Bare {{content}}? Sections with headings? AI summary? Custom formatting? read: the user's existing template files to learn how they capture things
4. Interpreter usage: Does the user want LLM-powered extraction (summaries, structured data, transformations)? Requires enabling Interpreter in the extension.
5. Target folder: Prefer the template folder the user already defined, and record it as user-template-folder/Clipper.

Anti-patterns