Skip to main content
Jam asks ChatGPT, Gemini, Perplexity, Google AI Mode, Google AI Overviews and Microsoft Copilot (and Claude, as an add-on) the questions your buyers ask. It records every answer and shows whether your brand is mentioned, recommended and cited. Research turns the gaps into a Plan: pages to write or fix, technical changes, sources to earn. Jam can do the work and ship it to your site, or hand it to your own agent.

Quickstart

Set up your site in the dashboard and read your first results.

Onboard with your agent

Connect Claude, Claude Code, ChatGPT, Cursor or any MCP client.

How Jam works

1

Tracking

Jam asks every tracked prompt on every platform your plan covers, on your plan’s cadence, and keeps each answer. See Tracking.
2

Research

Jam reads the answers to learn which pages AI assistants trust, which sources cite your competitors and where you are missing. See Research.
3

Actions

Research becomes the Plan: a ranked list of actions, each with its evidence. Send one to Jam, copy its prompt into your own agent, or do it yourself. See The Plan.
4

Delivery

Finished pages reach your site as a pull request on GitHub, or as Markdown you publish yourself. See Delivery.

Two ways to work

  • The dashboard at spreadjam.com. Everything in these docs has a button there.
  • Your own agent. The MCP server at https://api.spreadjam.com/mcp and the REST API at https://api.spreadjam.com/v1 expose everything the dashboard does. Claude and ChatGPT sign in with OAuth; other clients send an API key.
Work that writes pages, opens pull requests or runs research again spends credits. Scans of your tracked prompts do not. See Plans and credits.