MANAGE ECLIPSE FROM YOUR FAVOURITE AGENT

Connect your account to Claude Code, Cursor, Codex or Claude Desktop, then ask for what you need. Check what moved this week, draft an article, research a competitor, schedule a week of posts. It all runs on the same permissions and billing as the app.

  • Claude Code
  • Cursor
  • Codex
  • Claude Desktop

Your agent can reach the whole platform

Anything you would open Eclipse to do, you can ask for instead.

Ask what moved this week

Keyword ranks and history, visibility across ChatGPT, Claude, Gemini and Perplexity, traffic and search performance from GA4 and Search Console, and whatever the monitors picked up.

Draft the article and the posts from it

Write and edit against the client's brand profile, build a Cortex canvas from URLs and videos, research a competitor, then turn the finished article into social drafts.

Publish it where it needs to go

Send an approved article to WordPress, Webflow, HubSpot, Contentful, Wix, Ghost or Google Docs, and schedule posts through Buffer or a direct social connection.

The kind of thing people ask it

  • “Which keywords moved most in the last 30 days, and which pages are they on?”
  • “Turn yesterday’s article about operations into three LinkedIn posts and schedule them for next week.”
  • “Build me a canvas from these five competitor pages and this conference talk.”
  • “Where does our visibility in AI search lag behind our competitors?”
  • “Which of my clients has the most drafts waiting for review?”
  • “Every Monday, tell me which keywords moved and draft three posts from the top article.”

Common questions

What teams ask before they connect an agent to their account.

What is MCP?

MCP is the open standard AI tools use to connect to other products. Eclipse hosts a server at one address, and your agent connects to it over HTTPS with a token you create in settings. There is nothing to download.

Which tools does it work with?

Claude Code, Cursor and Codex connect natively. Claude Desktop connects through a one-line bridge. Settings gives you the token and the config snippet for whichever one you use.

What can the agent see?

Exactly what you can see. Every request checks your rights the way the app does, so a direct customer reaches their own account and an agency user reaches the clients they are assigned to. A token can be bound to a single client, which is what you want for a contractor. Integration credentials are never returned.

Can it spend money or publish without asking?

No. Anything that spends credits or AI balance, posts externally or buys something comes back as a preview first, with the cost where the price is known. It only runs once you confirm. If the connection drops halfway through, retrying returns the original result rather than charging twice.

How long do things take?

Reads come back immediately. Research runs and visibility analyses hand back a job the agent follows, and most finish in minutes. Drafting social posts can take up to ten minutes, and the agent will offer to check back rather than sit there.

Is there a record of what it did?

Every call is recorded against the token, and credits and AI balance spent through an agent appear in the client's ledger against the person whose token was used, exactly as they would from the app.

What stays in the web app?

Subscriptions, invoices, users and roles, along with connecting and disconnecting integrations. Those are the decisions about what Eclipse can reach, and most of them are browser sign-in flows.

Connect your agent to Eclipse

Bring a client you already run in Eclipse and we will set the token up with you on the call.