#Integration This reference documents the MCP (Model Context Protocol) tools provided by Scraps for AI assistant integration. search_scraps Search through your Scraps content with natural language queries using fuzzy matching. Parameters: query (string, required): Search query to match against scrap titles and contexts num (integer, optional): Maximum number of results to return (default: 100) Returns: results: Array of matching scraps with their title, context, and full Markdown content count: Total number of matches found list_tags List all available tags in your Scraps repository with their backlink counts, sorted by popularity. Parameters: None Returns: Array of tags with the following fields: title: Tag name backlinks_count: Number of scraps referencing this tag lookup_scrap_links Find outbound wiki links from a specific scrap. Returns all scraps that the specified scrap links to. Parameters: title (string, required): Title of the scrap to get links for ctx (string, optional): Context if the scrap has one Returns: Array of linked scraps with their full content. lookup_scrap_backlinks Find inbound wiki links (backlinks) to a specific scrap. Returns all scraps that link to the specified scrap. Parameters: title (string, required): Title of the scrap to get backlinks for ctx (string, optional): Context if the scrap has one Returns: Array of scraps that link to the specified scrap, with their full content. lookup_tag_backlinks Find all scraps that reference a specific tag. Parameters: tag (string, required): Tag name to get backlinks for Returns: Array of scraps that reference the specified tag, with their full content. Notes All search and lookup operations are performed against the current state of your Scraps repository Fuzzy matching is used for search queries to improve discoverability Results include the full Markdown content of matching scraps The MCP server must be running for these tools to be available to your AI assistant For setup instructions, see Integrate with AI Assistants.

#Integration Currently, the most recommended LSP for editing Scraps is markdown-oxide. markdown-oxide supports the following editing environments: Neovim VSCode Zed Helix To match the current features provided by Scraps, place the following configuration file .moxide.toml under the scraps/ directory and open the scraps/ directory directly for a comfortable editing experience. heading_completions = false title_headings = false tags_in_codeblocks = false references_in_codeblocks = false We are considering a feature to generate the LSP configuration file during the init command.

#Integration Scraps includes comprehensive Model Context Protocol (MCP) server functionality, enabling AI assistants to directly interact with your Scraps knowledge base. What is MCP? The Model Context Protocol (MCP) is an open standard that enables AI assistants to securely access external data sources and tools. Scraps implements an MCP server that exposes your documentation as a searchable, linkable knowledge base. Quick Start Claude Code (Recommended) For Claude Code users, we provide an official plugin for seamless integration. See Install Claude Code Plugin for installation instructions. Manual MCP Server Setup For other MCP-compatible clients or advanced configurations, you can add Scraps as an MCP server directly: claude mcp add scraps -- scraps mcp serve --path ~/path/to/your/scraps/project/ Replace ~/path/to/your/scraps/project/ with the actual path to your Scraps project directory. For command details, see MCP Serve. Available Tools For detailed MCP tool documentation, see MCP Tools.

#Integration This guide shows you how to enable the Scraps MCP (Model Context Protocol) server plugin in Claude Code. Installation Step 1: Add the Plugin Marketplace First, add the Scraps plugin marketplace: claude plugin marketplace add boykush/scraps This registers the Scraps plugin catalog with Claude Code. Step 2: Enable the Plugin Add the following to your project’s .claude/settings.json: { "enabledPlugins": { "mcp-server@scraps-claude-code-plugins": true } } The plugin will automatically use the current directory as your Scraps project path. Configuration Custom Project Path (Optional) To specify a different Scraps project path, set the SCRAPS_PROJECT_PATH environment variable: { "env": { "SCRAPS_PROJECT_PATH": "/path/to/your/scraps/project" }, "enabledPlugins": { "mcp-server@scraps-claude-code-plugins": true } }