| name | supadata-mcp-research |
| description | Use when extracting transcripts or scraping/crawling web, video, or social content through Supadata MCP, including YouTube, TikTok, Instagram, Twitter/X, web pages, site maps, and creator/content research that needs structured external data. |
Supadata MCP Research
Use Supadata when a task needs fresh structured content from public web, video, or social sources.
Setup Notes
Supadata MCP docs: https://docs.supadata.ai/integrations/mcp
Claude Code user-scope MCP:
claude mcp add --transport http --scope user supadata https://api.supadata.ai/mcp
After adding, run /mcp in Claude Code and authenticate Supadata if prompted.
For local stdio MCP clients, Supadata can also run with:
env X_API_KEY=your-api-key npx -y @supadata/mcp
Never commit API keys. Store them only in the client MCP config, environment, or secret manager.
When To Use
- Pulling YouTube transcripts.
- Grounding scripts, hooks, and creator voice in real videos.
- Scraping public web pages for research.
- Crawling or mapping a site for context.
- Pulling structured data from supported social/video platforms.
- Comparing creator/channel/video patterns.
Workflow
- Identify the source URL, channel, site, or platform.
- Use Supadata MCP tools when available before falling back to manual transcript paste or generic web scraping.
- Store only the useful extracted facts, timestamps, and source URLs.
- For content creation, extract patterns rather than copying source language.
- For research, preserve source links and note retrieval date when freshness matters.
- If Supadata cannot fetch a source, report that and use the next-best source.
Output Rules
- Do not claim transcript access unless Supadata or another tool actually returned it.
- Keep direct quotes short.
- Include source URLs.
- Separate extracted facts from interpretation.
- Respect platform terms and privacy boundaries.
- Avoid storing raw large transcripts in skills; use summaries or source references.
Skill Stacking
Pair with:
youtube-transcript-researcher for creator/style analysis.readwise-content-miner for second-brain idea mining.plugin-research-workflow for multi-source research.artifact-pipeline to turn extracted material into scripts, briefs, diagrams, or decks.
Source Notes
Created from the Supadata MCP docs and user-provided MCP request. Supadata's documented MCP features include transcript extraction, web scraping, site mapping, and crawling across YouTube, TikTok, Instagram, Twitter/X, and web pages.
