Bring your archive
into your agent.

VaatChitra runs a remote MCP server. Connect it once and your AI assistant can read, search, export, and share your own transcripts — without you ever opening the dashboard.

Server Endpoint

https://vaatchitra.prashamhtrivedi.app/mcp

Authentication reuses your normal email one-time-code login over OAuth. Your client opens a browser the first time you connect; after that it stays signed in. Every tool is scoped strictly to your own recordings.

What your agent can do

Browse the archive

List and search every recording — filter by language, status, date, or words inside the transcript.

Read transcripts

Pull a full transcript with speaker labels and timestamps, original alongside translation.

Export inline

Hand any transcript back as readable Markdown or structured JSON, right inside the chat.

Track processing

Ask whether a freshly uploaded recording is still transcribing, translating, or done.

Share in one breath

Mint a 30-day public share link for a finished transcript without leaving your agent.

Connect in under a minute

CLI

Claude Code

Run one command, then approve the browser login when it pops up.

claude mcp add --transport http vaatchitra https://vaatchitra.prashamhtrivedi.app/mcp

Desktop

Claude Desktop

Open Settings → Connectors → Add custom connector, name it VaatChitra, and paste the server URL:

https://vaatchitra.prashamhtrivedi.app/mcp

Editor

Cursor

Add this to ~/.cursor/mcp.json (or the project-local .cursor/mcp.json), then reload:

{
  "mcpServers": {
    "vaatchitra": {
      "url": "https://vaatchitra.prashamhtrivedi.app/mcp"
    }
  }
}

Any

Other clients

VaatChitra speaks the standard Streamable HTTP transport with OAuth. Any MCP-aware client takes the same URL:

https://vaatchitra.prashamhtrivedi.app/mcp

If your client asks for a config block instead of a URL, use the Cursor JSON above.

Need an archive to connect to?

Upload a recording first — your agent can only read transcripts that already live in your account.

Open VaatChitra