Summarize a web article and create a scrap with source link and tags.
View on GitHubplugins/scraps-writer/skills/web-to-scrap/SKILL.md
February 5, 2026
Select agents to install to:
npx add-skill https://github.com/boykush/scraps/blob/main/plugins/scraps-writer/skills/web-to-scrap/SKILL.md -a claude-code --skill web-to-scrapInstallation paths:
.claude/skills/web-to-scrap/# Web to Scrap Summarize a web article and create a scrap with Wiki-link notation. ## Arguments - **url**: `$ARGUMENTS` - URL of the web article to summarize - **max-lines**: (optional) - Maximum number of lines for the generated scrap ## Workflow 1. **Fetch the Article** - Use `WebFetch` to retrieve the web article content - Extract the title and main content 2. **Research Existing Tags** - Use `list_tags` to get available tags - Identify tags relevant to the article topic 3. **Search Related Scraps** - Use `search_scraps` to find related content - **NEVER create Wiki-links that were not returned by `search_scraps`. If a scrap is not in the results, it does not exist.** 4. **Create the Scrap** - **Only use Wiki-links to scraps found in step 3. Do not link to anything else.** - Generate a concise summary of the article - Include the source URL as autolink: `<https://...>` - Write well-structured Markdown content following the syntax below ## Wiki-Link Syntax - `[[Page Name]]` - Normal link. Use `title` from `search_scraps` results. - `[[Page Name|Display Text]]` - Alias link with custom display text. - `[[Context/Page Name]]` - Context link. Use `ctx/title` from `search_scraps` results. - `#[[Tag Name]]` - Tag. Use tags from `list_tags` results only. ## Markdown Features - CommonMark specification - GitHub-flavored Markdown (tables, task lists, strikethrough) - Mermaid diagrams with `mermaid` code blocks - Autolinks for OGP cards: `<https://example.com>`