BrainTube
BrainTube is an MCP server that semantically searches saved YouTube videos, articles, podcasts, PDFs, Notion pages, and Obsidian notes so developers can query personal research from inside their AI editor.
About
BrainTube MCP turns a developer's saved learning material into an agent-queryable knowledge base accessible from Claude Code, Cursor, Windsurf, and other MCP hosts. Connect via the hosted MCP endpoint at braintube-mcp-production.up.railway.app using a Bearer JWT API key issued from brain-tube.com Settings. Six MCP tools—search, smart_resurface, get_knowledge_graph, brain_chat, list_brains, and get_item—cover semantic lookup across brains, spaced-repetition review queues, knowledge-graph connections, conversational Q&A per brain, brain inventory with stats, and full item enrichment payloads. Indexed content types include YouTube videos, web articles, podcasts, PDFs, Notion pages, and Obsidian notes synced via the BrainTube Obsidian community plugin. Developers reach for BrainTube when a coding agent must cite or resurface material they already captured—not for live web crawling, git-repo indexing, or spinning up a self-hosted vector stack.
- Exposes 6 MCP tools: search, brain_chat, get_item, list_brains, smart_resurface, get_knowledge_graph
- Indexes 6 source types: YouTube, articles, podcasts, PDFs, Notion, and Obsidian
- JWT Bearer auth on hosted endpoint braintube-mcp-production.up.railway.app/mcp
- brain_chat enables conversational Q&A against a named brain collection
- smart_resurface returns spaced-repetition review queue items due for review
BrainTube by the numbers
- Data as of Jul 7, 2026 (Skillselion catalog sync)
claude mcp add --transport http braintube-mcp https://mcp.brain-tube.com/mcp --header "X-BrainTube-Token: YOUR_X_BRAINTUBE_TOKEN"Add your badge
Show developers this MCP server is listed on Skillselion. Paste this into your README.
| Transport | HTTP |
|---|---|
| Auth | Required |
How do you search saved YouTube and notes via MCP?
Run JWT-authenticated semantic search, brain_chat, and get_item over saved YouTube, PDF, Notion, and Obsidian content from Claude or Cursor.
Who is it for?
Developers who save research across YouTube, articles, PDFs, Notion, and Obsidian and want RAG-style retrieval inside Claude or Cursor.
Skip if: Developers needing live web crawling, codebase indexing, or a self-hosted-only knowledge stack without the BrainTube cloud service.
What you get
Ranked search hits, brain chat answers, knowledge-graph connections, and enriched item metadata retrieved from your BrainTube knowledge base inside the MCP host.
- Semantic search result sets
- Brain chat responses with source items
- Enriched item metadata and knowledge-graph connections
By the numbers
- Exposes 6 MCP tools: search, smart_resurface, get_knowledge_graph, brain_chat, list_brains, get_item
- Indexes 6 content source types: YouTube, articles, podcasts, PDFs, Notion, and Obsidian
Recommended MCP Servers
America Law GraphSearch 529K US statute sections.
Bakom McpBAKOM telecommunications and media open data
Changelog Ai Mcpchangelog-ai-mcp MCP server by MEOK AI LabsCron Ai Mcpcron-ai-mcp MCP server by MEOK AI LabsFlight Logger Mcpflight-logger-mcp MCP server by MEOK AI Labs
Foundation DiscoveryFoundation discovery and grant intelligence for nonprofits.How it compares
Pick BrainTube when research already lives in saved videos and notes; pick codebase RAG or web-crawl MCPs when the target is repository source or live pages.
FAQ
What content sources does BrainTube MCP index?
BrainTube indexes YouTube videos, web articles, podcasts, PDFs, Notion pages, and Obsidian notes. Obsidian vaults sync via the BrainTube community plugin with per-brain markdown folders and frontmatter metadata.
How do you authenticate BrainTube MCP requests?
BrainTube MCP requires a JWT API key passed as an Authorization Bearer header on every MCP and REST request. Keys are generated at brain-tube.com under Settings → API Keys.
Which MCP tools does BrainTube expose?
BrainTube exposes search, smart_resurface, get_knowledge_graph, brain_chat, list_brains, and get_item—covering semantic lookup, spaced-repetition review, graph traversal, conversational retrieval, and full item enrichment.