Nooticr — Claude Code MCP server: what it does, how to install
Pulls social posts, video frames, and comments across 10 platforms into your assistant. Useful for competitive research and brand monitoring.
claude mcp add --transport http mcp https://mcp.nooticr.com/mcpWhat Nooticr does
This MCP server exposes 169 tools to fetch social data across ten networks, including TikTok, YouTube, Reddit, and Instagram. It delivers uninterpreted material directly: post metrics, comment threads, up to 30 minutes of transcribed audio, and frame samples extracted via ffmpeg. The server does not process content with internal models; it hands raw data to your assistant.
Brand tracking covers both text and audio. The search_mentions tool looks for brand names inside comments across nine networks within a selected timeframe. For TikTok, Douyin, and YouTube, search_spoken_mentions identifies words spoken on camera that never appeared in typed captions.
Calls consume Nooticr credits, with 20 free credits granted on new accounts. Visual context has an overhead: each retrieved video frame takes roughly 1,200 tokens in your model window.
What it helps with
- extract video frames and audio transcripts from TikTok or YouTube
- track brand name occurrences in comments across nine networks
- find spoken brand mentions inside video audio tracks
- classify post comments by sentiment, complaints, and questions
Example prompts for the agent
- Extract frames and the transcript from this video URL and break down the structure
- Find recent brand mentions in social comments from the past seven days using
search_mentions - Sort the comment section of this post by complaints and feature requests
How to install Nooticr
1Platform CLI: remote server
claude mcp add --transport http mcp https://mcp.nooticr.com/mcp2Platform CLI
claude mcp add nooticr -- npx -y @nooticr/mcp1Platform CLI: remote server
codex mcp add nooticr --url https://mcp.nooticr.com/mcp2Platform CLI
codex mcp add nooticr -- npx -y @nooticr/mcpConfig file
{
"mcpServers": {
"nooticr": {
"url": "https://mcp.nooticr.com/mcp"
}
}
}{
"mcpServers": {
"nooticr": {
"command": "npx",
"args": [
"-y",
"@nooticr/mcp"
]
}
}
}Config file
{
"servers": {
"nooticr": {
"type": "http",
"url": "https://mcp.nooticr.com/mcp"
}
}
}{
"servers": {
"nooticr": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@nooticr/mcp"
]
}
}
}1Platform CLI: remote server
gemini mcp add -t http nooticr https://mcp.nooticr.com/mcp2Platform CLI
gemini mcp add nooticr npx -y @nooticr/mcpStep by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Nooticr
How to install Nooticr in Claude Code?
To connect via HTTP, run claude mcp add --transport http nooticr https://mcp.nooticr.com/mcp. If you prefer local execution over stdio, run claude mcp add nooticr -- npx -y @nooticr/mcp.
How does billing work?
Actions spend Nooticr credits, and new accounts receive 20 complimentary credits. Fetching a single post costs 1 credit, while deeper extraction jobs typically cost 2 or 3 credits.
How are video frames handled?
The get_post_frames tool inspects media streams directly using ffmpeg and returns actual images. Each returned frame consumes around 1,200 tokens of context.
