Veedcrawl/mcp

Veedcrawl MCP Server is the official Model Context Protocol (MCP) server exposing Veedcrawl’s social-video discovery and analysis tools to MCP-compatible agents (Claude, Cursor, Windsurf, etc.). It lets agents search across YouTube, TikTok, Instagram, X/Twitter and Facebook without needing pre-collected URLs, then fetch normalized metadata, full transcripts, run structured extraction, and retrieve creator profile snapshots. Who uses it: developers building AI assistants, analysts and researchers doing competitive or content analysis, and automation that needs scalable video discovery. Key features: - search_social_videos: cross-platform topic/keyword/trend searches with normalized results - get_video_metadata: structured title/author/engagement/tags/duration - get_video_transcript: native, AI-generated, or auto fallback with timestamps - extract_from_video: freeform or JSON Schema–constrained structured extraction - get_tiktok_profile / get_instagram_profile: creator snapshots with recent posts and per-post stats - npm install @veedcrawl/mcp; uses VEEDCRAWL_API_KEY; integration examples for Claude, Cursor, etc.

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.