LABSCO SUMMARY
Server-side plans with branching and loops move the control flow out of the model, which is where it belongs.
An agent orchestrating a ten-step sequence spends a turn per step and can lose the thread halfway. Sending a typed plan that the server executes — with real branching and iteration — makes the sequence deterministic and cheap. Whether synthetic audiences answer your question is a separate judgement, and one worth making before designing around the output.
What it is
Neuroflash's endpoint for their synthetic-audience product. What distinguishes it is Plan-Mode: instead of the client orchestrating a sequence of calls, a typed plan with branching and loops runs server-side.
What you get
- twins.query — poll synthetic audiences and segments
- content.generate — branded content generation
- Three modes: traditional per-endpoint tools, a server-side typed Plan-Mode supporting branching and loops, and an exploratory beta mode
- 88 tools in Traditional mode, 2 in Plan-Mode and 3 in the Exploratory beta
Requirements
A neuroflash account, signed in over OAuth by your client; one account covers the app, the API and MCP. Sign-in works in clients that return to a web address neuroflash accepts or to localhost — Claude, ChatGPT, Claude Code, Codex, VS Code and most others. Cursor is refused on a direct connection: connect it through the mcp-remote bridge.
Setup effort
Paste a URL, then authorize — add the endpoint to your client, then approve the OAuth consent screen
USE CASES
01
Polling a synthetic audience
Query the twins directly rather than fielding a real survey.
02
Running a multi-step job cheaply
Send a typed plan with its branching and let the server execute it.
03
Generating on-brand content
Content generation that uses the brand context already configured.
BEFORE YOU INSTALL
What to know first
Tools you get
88 tools by neuroflash's count.
Requires
- an MCP client that can reach a remote endpoint
- A neuroflash account, signed in over OAuth 2.0 with PKCE (Cursor through the mcp-remote bridge)
INSTALLATION
Connect your client
Add the endpoint to your client, then approve the OAuth consent screen.
Server setup
ENDPOINT URL
https://app.neuroflash.com/api/mcp-server/v1/mcpRequirements
- OAuth 2.0 with PKCE, signing in with your neuroflash account. A client completes the sign-in itself only when its redirect is an allowed HTTPS address or localhost; Cursor (cursor://) is refused and signs in through the mcp-remote bridge
MCP Clients
Claude Desktop
Open Settings (gear icon) and go to Connectors
Scroll down and click Add custom connector
Enter the server URL: https://app.neuroflash.com/api/mcp-server/v1/mcp and click Add
A browser window opens — log in with your neuroflash account to authorize the connection
Cursor
.cursor/mcp.json in your project root (or the global Cursor MCP config at ~/.cursor/mcp.json)
CURSOR · JSON
{
"mcpServers": {
"neuroflash": {
"command": "npx",
"args": [
"mcp-remote",
"https://app.neuroflash.com/api/mcp-server/v1/mcp"
]
}
}
}