Spendict
About
Four MCP tools, one performance marketer's judgment — design targeting, audit a campaign build, gate an ad creative before spend, and diagnose a live campaign. Deterministic, server-verified verdicts your agent can act on.
Details
- Author
- dinosaur24
- Downloads
- 331
- Categories
- Marketing, Other, AI
Jump to
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
SpendictCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
{
"mcpServers": {
"spendict": {
"type": "http",
"url": "https://www.spendict.com/api/mcp"
}
}
}
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"spendict": {
"spendict": {
"type": "http",
"url": "https://www.spendict.com/api/mcp"
}
}
}
}McpServers
{
"spendict": {
"type": "http",
"url": "https://www.spendict.com/api/mcp"
}
}What it is
Your AI agent can generate a hundred ad variants an hour — and has no idea which ones are any good, because the model that wrote the ad is the same model grading its own homework. Spendict is a performance marketer's judgment, shipped as an MCP server. Add it to Claude, Cursor, Windsurf, n8n, or any MCP client and four tools appear directly in your agent's tool list — no bespoke integration. A REST API mirrors every tool for callers that don't speak MCP yet.
A media buyer's whole job, in four tools
- strategize_targeting — design the plan
- audit_campaign_structure — check the build before budget flows
- assess_ad_creative — gate the creative before a cent is spent
- analyze_campaign_performance — diagnose the live numbers, and reconcile the pre-flight call against reality
The demoable moment: an agent generates 10 ad variants, calls assess_ad_creative on each, and only the ones that clear the bar (or get fixed first) get budget — the doomed ones are killed before spend.
Why trust the verdict
The server, not the model, decides. Every tool returns a structured verdict whose score and recommendation are recomputed deterministically in server code — a confident-but-wrong model can't talk its way past a hard gate. The judgment is a calibrated expert opinion, hand-tuned against real ads by a performance marketer who has actually spent on them.
Pricing & access
Works with Claude, Cursor, Windsurf, n8n, and any MCP client. Free for 50 checks/mo, Starter $19/mo (1,000), Pro $49/mo (5,000). Quota is checked before inference and failed calls are refunded — you're never charged for an error. Get a key at spendict.com.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
