About Lection
Lection is a hosted MCP server for searching public YouTube content and accessing a user's Lection extraction data. It combines tabular search results with tools to inspect saved runs and manage schedules.
Key features
- 01Search YouTube videos by keyword
- 02List a channel's latest uploads
- 03Retrieve public video comments
- 04View saved scrapes and run results
- 05Create and pause extraction schedules
Use cases
- 01Find videos matching a research topic
- 02Review recent uploads from a channel
- 03Retrieve comments from a public video
- 04Manage scheduled data exports
MCP support
Lection provides a remote MCP server at https://www.lection.app/mcp. Public YouTube tools need no sign-in; account tools use OAuth 2.1 to access the signed-in user's Lection data.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://www.lection.app/mcpclaude mcp add --transport http eteleieto-lection https://www.lection.app/mcp{"mcpServers":{"eteleieto-lection":{"type":"http","url":"https://www.lection.app/mcp"}}}Open in Cursor{"servers":{"eteleieto-lection":{"type":"http","url":"https://www.lection.app/mcp"}}}Open in VS Code{"mcpServers":{"eteleieto-lection":{"type":"http","url":"https://www.lection.app/mcp"}}}{"mcpServers":{"eteleieto-lection":{"serverUrl":"https://www.lection.app/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
11| Parameter | Type | Required | Description |
|---|---|---|---|
| query | string | Yes | Search keywords, exactly as you would type them into YouTube. |
| maxRows | integer | No | How many rows to return (1-20, default 20). The free ChatGPT app is capped at 20. |
Details
From the project README
Lection MCP server
A hosted (remote) Model Context Protocol server for Lection, a tool that turns web pages into structured tables and exports them to CSV, Excel, Google Sheets or a webhook on a schedule.
- Endpoint:
https://www.lection.app/mcp(Streamable HTTP) - Nothing to install, no API key for the public YouTube tools.
- Account tools use OAuth 2.1 (Supabase). Protected-resource metadata: https://www.lection.app/.well-known/oauth-protected-resource
Tools
Public (no sign-in, results capped at 20 rows, official YouTube Data API):
| Tool | What it does |
|---|---|
youtube_search | Search YouTube videos by keyword: title, channel, views, date, URL. |
youtube_channel_videos | Latest uploads of a channel by @handle or URL. |
youtube_comments | Top-level comments of a public video: author, text, likes, replies, date. |
lection_browser_tools_info | How to export Instagram comments/followers or LinkedIn connections with Lection (those sites need a login, so they run in the Lection browser extension). |
Account tools (OAuth; act only on the signed-in user's own Lection data):
| Tool | What it does |
|---|---|
my_lection_account | Plan, schedule limits, allowed destinations, cloud credits. |
list_my_saved_scrapes | Recent saved scrapes. |
get_run_results | Rows of a saved scrape or scheduled run as a table. |
list_my_schedules | Scheduled extractions with cadence, destination, next run, status. |
list_my_runs | Recent schedule runs with status and row counts. |
schedule_extraction | Create an hourly, daily or weekly schedule for a saved scrape (email, Google Sheet or webhook). |
pause_schedule | Pause a schedule (resumable in the dashboard). |
All read-only tools carry readOnlyHint: true; pause_schedule is marked destructive and asks for confirmation.
Connect
Claude (web, desktop, mobile)
Settings → Connectors → Add custom connector → URL https://www.lection.app/mcp → Add. Sign in only if you want the account tools.
Claude Code
claude mcp add --transport http lection https://www.lection.app/mcp
ChatGPT
Enable Developer mode (Settings → Apps → Advanced), then create an app / custom MCP server with URL https://www.lection.app/mcp. Authentication: none for the YouTube tools, OAuth for account tools.
Cursor / VS Code
Add to .cursor/mcp.json (Cursor) or .vscode/mcp.json (VS Code, under "servers"):
{
"mcpServers": {
"lection": {
"url": "https://www.lection.app/mcp"
}
}
}
stdio-only clients (via mcp-remote)
{
"mcpServers": {
"lection": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://www.lection.app/mcp"]
}
}
}
Example prompts
- "Show the top YouTube videos about n8n automation tutorials."
- "List the latest videos from @OpenAI."
- "Get the top comments on this video: https://www.youtube.com/watch?v=dQw4w9WgXcQ"
Privacy and terms
Privacy policy: https://www.lection.app/privacy · Terms: https://www.lection.app/terms · Support: [email protected]
YouTube results come from the YouTube Data API and are subject to the YouTube Terms of Service and Google Privacy Policy. Lection does not store them.
License
MIT (this repository's metadata and docs). The hosted service is operated by Lection.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.