JetBrains AI MCP setup: TikTok and Instagram data in AI Assistant and Junie

Updated

Virlo is social listening for TikTok and Instagram Reels. JetBrains AI Assistant and Junie understand your project deeply, yet neither can tell which TikTok or Instagram videos are outperforming this week. Connect Virlo over MCP and both can query trends, hooks, sounds and outlier videos from a database of 23.1M viral videos without leaving IntelliJ IDEA, PyCharm, WebStorm or any other JetBrains IDE. JetBrains AI is the AI layer in JetBrains IDEs: AI Assistant for chat and edits, and Junie, its autonomous coding agent.

  • AI Assistant: add Virlo in Settings | Tools | AI Assistant | Model Context Protocol (MCP)
  • Junie: one url entry in ~/.junie/mcp/mcp.json
  • Global or project level, so a team can share the setup

Part of Virlo's AI assistant guides. Using Copilot inside a JetBrains IDE instead? See GitHub Copilot. Prefer an extension agent? See Cline.

How do I add an MCP server to JetBrains AI Assistant and Junie?

AI Assistant and Junie keep separate MCP lists, so add Virlo to whichever you use, or both. Each takes a couple of minutes.

  1. Step 1: Create your Virlo account for free

    Create your free Virlo account in under a minute. Already signed up? Open API Keys.
  2. Step 2: Create a Virlo API key

    Generate a key on the Virlo API Keys page. It starts with virlo_tkn_ and replaces virlo_tkn_YOUR_KEY in the snippets below.

  3. Step 3: AI Assistant: add Virlo in the MCP settings

    Open Settings | Tools | AI Assistant | Model Context Protocol (MCP), click Add, paste this JSON into the New MCP Server dialog, click OK and then Apply. It starts the mcp-remote bridge with Node.js, which forwards your key to Virlo. Set the Server level to Global to use Virlo in every project.

    New MCP Server (JSON)
    {
      "mcpServers": {
        "virlo": {
          "command": "npx",
          "args": [
            "-y",
            "mcp-remote",
            "https://dev.virlo.ai/api/mcp/mcp",
            "--header",
            "Authorization: Bearer ${VIRLO_API_KEY}"
          ],
          "env": {
            "VIRLO_API_KEY": "virlo_tkn_YOUR_KEY"
          }
        }
      }
    }
  4. Step 4: Junie: add Virlo to mcp.json

    Open Junie's MCP settings, click Add, and the mcp.json file opens. Paste this entry to make Virlo global (the file lives at ~/.junie/mcp/mcp.json), or put it in .junie/mcp/mcp.json at the project root to share it with the repo. Junie also lets you drop the headers block and pick Authorize to sign in through the browser.

    ~/.junie/mcp/mcp.json
    {
      "mcpServers": {
        "virlo": {
          "url": "https://dev.virlo.ai/api/mcp/mcp",
          "headers": {
            "Authorization": "Bearer virlo_tkn_YOUR_KEY"
          }
        }
      }
    }
  5. Step 5: Ask AI Assistant or Junie

    Check that virlo shows as active in the MCP list, then open AI Chat or hand Junie a task and ask in plain English.

    Your first prompt
    Use Virlo to list the 10 TikTok trends to copy with the most creators in [your niche] this week and write them to research/trends.md with an example video for each.

The AI Assistant route needs Node.js on your machine so the IDE can run npx. Junie connects to Virlo by URL directly, with no bridge.

AI Assistant vs Junie vs Junie CLI: how does each JetBrains tool connect to Virlo?

JetBrains toolWhere Virlo is addedHow it authenticatesBest for
AI AssistantSettings | Tools | AI Assistant | Model Context Protocol (MCP), AddAPI key passed by the mcp-remote bridgeAsking questions and getting edits in AI Chat
Junie in the IDEJunie's MCP settings, Add, which opens mcp.jsonAuthorization header with your key, or Authorize for a browser sign-inMulti-step tasks that should write research files
Junie CLI~/.junie/mcp/mcp.json, the same file the IDE plugin usesSame as Junie in the IDERunning Junie from a terminal or CI
Project-level configAI Assistant Level column set to Project, or .junie/mcp/mcp.json in the repoEach developer brings their own key or sign-inTeams that want Virlo only in one codebase

Both tools reach the same Virlo server, so a trend list in AI Chat and one in a Junie task come from identical data.

Which hook types are winning on Instagram, according to what Junie can fetch?

A live table JetBrains AI can pull from Virlo: hook types ranked by how often Instagram Reels using them reach a Virality Score of 18 or more, set beside the TikTok rate for the same hook.

In Virlo's snapshot of 264,108 hook-tagged Instagram videos, comparison hooks reached a Virality Score of 18 or more 19.0% of the time, the best rate on Instagram.

As of November 19, 2026

Which hook types are winning on Instagram, according to what Junie can fetch?
Hook typeInstagram: share at 18+TikTok: share at 18+Instagram videos
Comparison19.0%29.8%2,947
POV setup18.1%32.9%8,409
Shock statement16.8%31.8%12,033
Trend reference15.3%30.3%1,106
Relatable scenario15.3%31.2%19,312
Before and after14.6%28.6%1,892

Hook types by niche, with the Reels behind them

Split by 20+ niches, with every hook and video behind each number, in the Virlo app.

Open hooks in Virlo

How this sample is made

Source
Virlo's daily hook snapshot: 264,108 Instagram and 1,365,774 TikTok videos whose opening hook was tagged by type.
Window
All tagged videos in the snapshot. Hook types with fewer than 1,000 videos on either platform, and videos with no hook, are left out.
Method
Share at 18+ = the share of videos with that hook type whose Virality Score is 18 or more. Sorted by the Instagram rate.
As of
November 19, 2026. Rebuilt every day.
Made by
Numbers are pulled automatically from Virlo's data; the text around them is written by the Virlo team.

Suggested citation: Source: Virlo, “JetBrains AI MCP setup: TikTok and Instagram data in AI Assistant and Junie”, virlo.ai/jetbrains

A JetBrains workflow for content research next to your code

Teams that build marketing sites, apps or internal tools in a JetBrains IDE can keep their social research in the same project. AI Assistant is quicker for one-off questions, and Junie suits jobs that end with several files written.

  • In AI Chat: "Which sounds are rising on Instagram Reels this week?" for a fast answer while you work.
  • In Junie: "Build research/weekly/[date].md with trends, hooks and sounds for [your niche]" as a full task.
  • Ask Junie to turn the report into a Kotlin, Python or TypeScript data class so the data can drive a page.
  • Keep the reports under version control so the team can compare weeks in the IDE's diff viewer.

Virality Score ranks how far a video outran its creator's normal views. Scores of 18 and up are strong, which surfaces breakout ideas from small accounts as well as big ones.

The JetBrains MCP server is a different thing

JetBrains IDEs can also act as an MCP server, letting outside agents such as Claude Code read and edit your project through the IDE. That feature points the other way: it gives agents access to the IDE, while this guide gives the IDE's own AI access to Virlo.

You can run both at once. Junie can even be auto-configured to talk to the IDE's MCP server, and Virlo sits alongside it in the same mcp.json.

Guidelines that steer Junie toward Virlo

Junie follows project guidelines kept in the .junie folder of your repo. A few lines there, such as "use the virlo MCP server for TikTok and Instagram questions, rank by Virality Score and link each source video", make every Junie task on that project treat Virlo as its social data source.

AI Assistant has its own project rules in the IDE settings. Adding the same instruction there keeps chat answers and Junie tasks consistent, which helps when one teammate asks in AI Chat and another hands Junie a full task on the same topic.

What can I ask JetBrains AI once Virlo is connected?

Use these in AI Chat or as Junie tasks. Replace the brackets with your own details.

  • Weekly trends

    Write research/trends.md with this week's top 10 TikTok trends to copy in [your niche], each with the part to keep and an example video.

  • Data class

    Fetch one Virlo trend record for [your niche] and generate a matching data class in [language] under [package path].

  • Reels hooks

    Which hook types reach a Virality Score of 18+ most often on Instagram Reels in [your niche]? Put the table in docs/hooks.md.

  • Competitor scan

    Look up @[competitor] on TikTok and list which videos beat their usual views, with the hook used in each.

  • Sound picks

    List five sounds trending on Instagram Reels this week and pick one that fits a [tone] video for [your product].

  • Transcript to script

    Get the transcript of [TikTok link] and turn it into a 30-second script for [your product] in scripts/draft.md.

  • Seed fixtures

    Create test fixtures from 15 real TikTok videos about [topic] pulled with Virlo, matching the schema in [file].

  • Creator shortlist

    Find 10 creators under [N] followers in [your niche] whose videos keep beating their follower count and save them as CSV.

JetBrains AI Assistant vs Junie vs the IDE MCP server: what is what?

JetBrains ships several AI pieces with overlapping names. Here is which one this guide configures.

JetBrains AI Assistant (this page)

The chat and in-editor helper. It takes Virlo through Settings | Tools | AI Assistant | Model Context Protocol (MCP).

Junie (this page)

JetBrains' autonomous coding agent, in the IDE and as a CLI. It reads Virlo from mcp.json.

JetBrains IDE as an MCP server

A built-in server that exposes your IDE to external agents. It does not add social data; Virlo does.

GitHub Copilot for JetBrains

GitHub's plugin for JetBrains IDEs, with its own MCP settings separate from AI Assistant and Junie.

What are the limits?

  • Virlo is read-only and covers public TikTok and Instagram data; JetBrains AI can study and write with it but cannot post or log in to accounts.
  • AI Assistant and Junie keep separate server lists, so a server added in one is not automatically present in the other.
  • On company licences, an administrator can control which AI features and MCP servers are allowed in the IDE.
  • Trends to copy refresh every hour and pass a check by the Virlo team.
  • For what each Virlo plan includes, see pricing.

Frequently asked questions

How do I add an MCP server to JetBrains AI Assistant?

Go to Settings | Tools | AI Assistant | Model Context Protocol (MCP), click Add, paste the server JSON in the New MCP Server dialog, then click OK and Apply. For Virlo, use the mcp-remote entry shown above, which passes your API key to the hosted server.

How do I add an MCP server to Junie?

Open Junie's MCP settings and click Add; the mcp.json file opens. Add a virlo entry under mcpServers with "url": "https://dev.virlo.ai/api/mcp/mcp" and an Authorization header holding your key. Save, and Junie lists Virlo as active. If Junie asks you to authorize instead, finish the browser sign-in and the status turns active.

Where is the Junie mcp.json file?

Global servers live in ~/.junie/mcp/mcp.json in your home directory, and project servers in .junie/mcp/mcp.json at the project root. Junie in the IDE and the Junie CLI read the same format, so one entry serves both. Commit the project file only if it holds no key, or rely on Authorize for each developer.

Does JetBrains AI Assistant support remote MCP servers?

Yes, it supports Streamable HTTP servers as well as local STDIO ones and legacy SSE. For Virlo the bridge route is the simplest way to send an API key header, and it works in every JetBrains IDE that has AI Assistant enabled.

Can JetBrains AI see TikTok and Instagram data?

With Virlo connected, yes. AI Assistant and Junie receive trends to copy, outlier videos ranked by Virality Score, hooks, sounds, hashtags, creators and TikTok transcripts as text, ready to summarise or write into project files. Every item links back to a real video, so the IDE can show where each idea came from.

Which JetBrains IDEs work with this setup?

Any IDE where AI Assistant or Junie runs, including IntelliJ IDEA, PyCharm, WebStorm, GoLand, PhpStorm, Rider and RubyMine. The MCP settings sit in the same place in each, so the steps on this page carry over unchanged. Restart the IDE after adding Virlo if the server does not appear straight away.

Junie shows Virlo as "Authorization required". What do I do?

Select the virlo server in Junie's MCP list, choose Authorize and finish the sign-in in the browser window that opens. Once the status changes to Active, Junie can call Virlo. Adding an Authorization header with your key avoids the prompt altogether.

Can Junie post to TikTok through Virlo?

No. Virlo never logs in to social accounts and offers no posting tools. Junie can research, rank and write with Virlo data inside your project, and publishing happens in your social apps or scheduler. Junie does not need any social login for research tasks.

Can I use AI Assistant and Junie with Virlo at the same time?

Yes. They keep separate MCP lists, so add Virlo to each one you use. AI Assistant then answers quick questions in AI Chat with live data, while Junie can run longer tasks that write reports or code from the same Virlo account and the same key.

Give your JetBrains IDE live social data

Add Virlo to AI Assistant or Junie and ask what your niche is copying on TikTok and Instagram this week.