Verant

Find spelling, grammar, and placeholder errors on live web pages and sites, with a fix for each.

Hosted MCP Server

npx add-mcp 'https://verant.ai/mcp'

Installs into Claude Code, Codex, Cursor and more

Documentation

Claude & Cursor

Proofread from Claude and Cursor

Verant runs a Model Context Protocol (MCP) server, so any agent that speaks MCP — Claude, Cursor, and the rest — can proofread a page or crawl a whole site without leaving the tool you already work in.

Website address

svg]:px-3 h-8 w-8 shrink-0 rounded-full p-0" type="submit" aria-label="Start proofreading">

Single Page Full Site

p]:text-xs!">

$ claude mcp add verant https://verant.ai/mcp

one connection · three tools · any agent

[ THE SHORT ANSWER ]

Verant exposes a Model Context Protocol (MCP) server at verant.ai/mcp, listed in the official MCP registry as ai.verant/verant. Connect it once over OAuth from any MCP client — Claude, Cursor, or your own agent — and the agent gets three tools: proofread_url for a single page, proofread_site for a full crawl, and get_scan for a scan's status and results.

Sign in to connect your Verant account. A free account gets 2 pages a month, shared between the web app and MCP. Both use the same proofreading engine and the same account allowance. Pro and Ultra accounts can also scan a whole site.

What the MCP server does

Use proofread_url to check one published page. Use proofread_site on Pro or Ultra to crawl a public site, up to your page limit. Both return a scanId at once. Call get_scan with that scanId every 10 seconds until the scan completes or fails. It returns the same findings as your report, with dismissed findings removed.

Each correction quotes your exact text, and a suggestion is returned only when the engine can defend it. The agent receives a short, precise list — not a wall of raw suggestions it has to sort.

Why proofread from your agent

You are already drafting, editing, and shipping copy inside Claude or Cursor. With Verant connected, you can ask the agent to check the page you just changed, or crawl the whole site before a release, and get the answer in the same thread — no tab-switch, no paste, no separate dashboard trip.

Because weak suggestions are dropped before the list comes back, an agent can act on it directly: surface the real issues, draft the fixes for your review, and move on. The same gate that keeps the list short for a person is what makes it safe to hand an agent.

Built for agents, safely

MCP requires sign-in through OAuth. Each scan uses your account allowance, including the 2 free pages a month. Each page checked uses 1 page. The agent can read only scans owned by your account.

The verbatim rule still holds over MCP: Verant flags your exact text and never edits your site. What the agent gets back is a list of corrections to act on, not changes already made.

Connect Verant to your AI client

01 Add Verant's MCP server in your client and point it at https://verant.ai/mcp (Streamable HTTP). It is listed in the official MCP registry as ai.verant/verant.02 Sign in to your Verant account when the client asks. A free account gets 2 pages a month, shared with the web app.03 Ask your agent to call proofread_url for one page or proofread_site for a full crawl on Pro or Ultra. Then poll get_scan with the scanId every 10 seconds. Open reportUrl for the full report.

yoursite.com

one URL · nothing to install

[ report · yoursite.com ]

3 finds · quoted verbatim · fix beside each

share link csv

you act on it · nothing changes your site

[ report · yoursite.com ]

3 finds · quoted verbatim · fix beside each

share link csv

you act on it · nothing changes your site

[ VERIFIED ]

What ‘verified’ means

Weak suggestions are dropped before they reach you. What remains quotes the exact text on your page, with the fix beside it.

[ EXPLORE ]

Explore Verant

Related reading: the full website proofreading pass , automatic weekly monitoring , and transparent, self-serve pricing .

[ FAQ ]

Frequently asked questions

Which AI tools can use Verant?

Any client that supports the Model Context Protocol — Claude, Cursor, and other MCP-capable agents. Verant's server is Streamable HTTP at verant.ai/mcp and is listed in the official MCP registry as ai.verant/verant.

Do I need a paid plan to use the MCP server?

No. Sign-in is required, and a free account gets 2 pages a month on web and MCP combined. Each page checked uses 1 page of your allowance. Full-site scans need Pro or Ultra.

What can the agent actually do?

Three tools: proofread_url starts a single-page scan; proofread_site starts a full-site scan on Pro or Ultra. Both return a scanId at once. Poll get_scan with it every 10 seconds to read progress and findings, then open reportUrl for the report.

Does the MCP server apply the same precision bar?

Yes — the MCP server runs the same engine as the web app, so the same precision bar applies: a correction is returned only when the engine can defend it. Your agent receives a short list, with your exact text quoted, not raw suggestions.

Does the agent change my site?

No. Verant flags your exact text and never edits or auto-applies anything. The agent gets back a list of corrections to act on; any change is yours to make.

See what Verant finds on your site. Free trial, no credit card.

svg]:px-4" href="/#scan"> Try it on your site