Skip to main content
Glama
ericx0

TikRSS MCP Server

by ericx0

TikRSS CLI & Official Model Context Protocol (MCP) Server

MIT License MCP Protocol Compliant TypeScript

The official command-line interface (CLI) and Model Context Protocol (MCP) Server for TikRSS — the high-signal daily briefing engine that distills hours of video & podcast content from creators you actually trust into actionable intelligence.

This unified package connects your local developer environment and frontier AI models (Claude Desktop, Cursor, Windsurf, and Google Antigravity) directly to your synthesized TikRSS knowledge streams.


✨ Features

  • 🤖 Native AI Agent MCP Server: Zero-setup integration with Claude Desktop and Cursor. Let your AI agents autonomously query creator briefings, search cross-video arguments, and fetch clean transcripts.

  • ⚡ Developer Terminal CLI: Read your daily video distillation stream right in your terminal with beautiful ANSI formatting, or export as clean Markdown to your local Obsidian vault.

  • 🎯 1-Click Fast Ingestion: Subscribe to YouTube channels (@handle), TikTok creators, or podcast feeds with a single command.

  • 🔍 Semantic Claims Search: Search across thousands of synthesized arguments and citations without watching 3-hour long podcasts.

  • 🛡️ Built for Power Tier: Full access to dedicated priority Whisper ASR transcription pipelines and unlimited creators.


Related MCP server: OmniContext

🚀 Quick Start (CLI)

1. Install & Login

You can install globally or run via npx:

# Install globally
npm install -g tikrss

# Or run directly with npx
npx -y tikrss whoami

Configure your TikRSS API Key (obtainable from your TikRSS Account Settings):

tikrss login tr_live_your_api_key_here

Your credentials are securely saved in ~/.config/tikrss/config.json with strict 0600 permissions. You can also supply the key via the TIKRSS_API_KEY environment variable.

2. Common CLI Commands

Fetch Today's Briefing

# Formatted terminal card output
tikrss today

# Export as clean Markdown
tikrss today --format markdown

# Raw JSON output for shell pipelines / jq
tikrss today --format json

Subscribe a Creator or Channel

tikrss add https://www.youtube.com/@karpathy
tikrss add https://www.youtube.com/@lexfridman

Search Claims & Topics

tikrss search "test-time compute scaling"

Export Video Distillation to Obsidian / Local Markdown

tikrss export karpathy-cot-2026 --out ~/Documents/ObsidianVault/Briefings

🤖 Model Context Protocol (MCP) Integration

TikRSS provides a native, compliant MCP Server over stdio. It connects your AI agents directly to your video digest archive.

1. Claude Desktop Setup

Open your Claude Desktop configuration file:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json

  • Windows: %APPDATA%\Claude\claude_desktop_config.json

Add the tikrss server:

{
  "mcpServers": {
    "tikrss": {
      "command": "npx",
      "args": ["-y", "tikrss", "mcp"],
      "env": {
        "TIKRSS_API_KEY": "tr_live_••••••••"
      }
    }
  }
}

Restart Claude Desktop. You will see the hammer icon with 5 TikRSS tools enabled!

2. Cursor Setup

  1. Open Cursor Settings > Features > MCP.

  2. Click + Add New MCP Server.

  3. Fill in:

    • Name: tikrss

    • Type: command

    • Command: npx -y tikrss mcp

    • Environment Variables: TIKRSS_API_KEY=tr_live_••••••••


🛠️ MCP Tools Reference

When connected via MCP, your AI assistant gains access to the following 5 tools:

Tool

Parameters

Description

tikrss_today

format (optional: markdown or json)

Retrieves today's synthesized briefings from your subscribed creators.

tikrss_read_video

video_id (required)

Reads full distillation (Executive Summary, Why It Matters, core claims, quotes, and research papers).

tikrss_search_claims

query (required)

Semantic search across all video briefings, arguments, and topics in your archive.

tikrss_subscribe

url (required)

Autonomously subscribes a creator URL to your daily morning briefing pipeline.

tikrss_get_transcript

video_id (required)

Retrieves cleaned full-text transcript with commercial sponsors and fluff stripped out.


💬 Example Prompts for AI Agents

Once configured, simply chat naturally in Claude or Cursor:

  • "What did Andrej Karpathy and Lex Fridman say about test-time compute in their latest videos?" (Claude automatically calls tikrss_search_claims)

  • "Give me today's TikRSS morning briefing and draft a 3-bullet summary for our team." (Claude automatically calls tikrss_today)

  • "Subscribe this YouTube channel: https://www.youtube.com/@hubermanlab to my daily TikRSS digests." (Claude automatically calls tikrss_subscribe)


⚙️ Environment Variables

Variable

Description

Default

TIKRSS_API_KEY

Your TikRSS API Key (starts with tr_live_)

None

TIKRSS_API_URL

Self-hosted or custom TikRSS API endpoint

https://tikrss.app

TIKRSS_CONFIG_DIR

Custom directory for config file

~/.config/tikrss


🔒 Security & Privacy

  • Minimal Permissions: The MCP Server runs strictly in stdio mode and only communicates with the official TikRSS HTTPS API endpoint.

  • Safe Token Storage: Local config files are written with POSIX file permissions 0600 (readable only by the current user).

  • No Telemetry: No third-party trackers, analytics, or background telemetry.


📄 License

MIT License © 2026 TikRSS Team

Related MCP Connectors

Related MCP Servers

  • A
    license
    Not graded
    quality
    D
    maintenance
    Enables AI agents to fetch and digest content from 30+ platforms (Twitter, YouTube, Reddit, etc.) via a unified API. Supports multi-format output, transcription, and direct Obsidian sync.
    18
    BSD 2-Clause "Simplified"
  • A
    license
    A
    quality
    B
    maintenance
    Connects Claude Desktop to Notion and YouTube transcripts with one-click installation, enabling natural language queries about your pages and video content.
    5
    2
    MIT
  • A
    license
    A
    quality
    B
    maintenance
    Let AI agents watch videos: local transcripts, speaker labels, scenes, chapters and exact-moment search from any video URL or file. Fully local, no API keys.
    4
    2
    AGPL 3.0
  • A
    license
    Not graded
    quality
    B
    maintenance
    Provides Claude Code with local-first multimodal semantic search across documents, code, images, and video, plus multi-engine web research with active counter-evidence hunting and source-verified, extract-only briefings.
    4
    AGPL 3.0