About Agent Ready by WorkPods
Agent Ready by WorkPods is a local MCP server and CLI for checking SEO, AI-answer visibility, and whether websites can be used by AI agents. It supports browser-based checks through its extension and runs page, site-file, and crawl checks in Node.js.
Key features
- 01Page and site SEO audits
- 02AI-answer visibility checks
- 03Agent accessibility inspection
- 04Keyword and backlink research
- 05Browser-based agent journeys
Use cases
- 01Audit a page and crawl a site
- 02Inspect AI answers for brand mentions
- 03Check whether agents can use page controls
- 04Test an agent task in a scratch tab
MCP support
The local MCP server connects to AI clients over stdio and returns check results. Browser-dependent tools use the Agent Ready extension and its local bridge.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
12Saved tool list · updated 4d ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| url | string | Yes | The page to check, an absolute http or https URL such as https://example.com/. |
| brandName | string | No | The site's brand name as people search for it, such as "Acme Running". When given, the organization identity check compares the home page schema with it. |
Details
From the project README@workpods/agent-ready
Local MCP server and CLI for the Agent Ready browser extension: check SEO, AI answer visibility (GEO), and whether AI agents can use a website (AXO).
Install the extension
Install Agent Ready from the Chrome Web Store.
The extension provides rendered-page checks, reads from open AI answer tabs, and agent journeys. Page, site-file, and crawl checks can also run in Node without the extension.
Quick start
Requires Node.js 22 or later. Print the installed package version:
npx -y @workpods/agent-ready --version
Check one page and its site files:
npx -y @workpods/agent-ready check https://example.com --fail-on high
The command prints a Markdown report. It exits with code 1 if a check fails or partly fails at
high or critical severity. Checks that need a browser are reported as not applicable. Running
npx -y @workpods/agent-ready without a subcommand prints usage and exits with code 2.
Use from an AI agent
In the extension's Settings → Agent connection, enable Connect to the local bridge when it runs. Keep Chrome open while the agent works.
For Claude Code:
claude mcp add agent-ready -- npx -y @workpods/agent-ready mcp
For Cursor (~/.cursor/mcp.json) or Claude Desktop (claude_desktop_config.json), add this to
the client's MCP settings:
{
"mcpServers": {
"agent-ready": {
"command": "npx",
"args": ["-y", "@workpods/agent-ready", "mcp"]
}
}
}
The client starts the local server over stdio. See agent setup for Codex, skills, and Firefox origin configuration.
What it can do
- Page and site checks: audit a page, crawl a site, and inspect robots.txt, sitemaps, llms.txt, and agent manifests.
- Agent visibility: inspect rendered text, accessible controls, and WebMCP tools; compare two pages. Requires the extension.
- AI answers: read an answer already open in a supported AI assistant tab and report brand mentions and citations. Requires the extension; it does not send a prompt.
- Research: find keyword ideas, keyword gaps, backlinks, and link gaps through DataForSEO. Requires the extension and your own provider account; usage may incur charges.
- Journeys: observe a page and take steps in a scratch tab to test an agent task. Requires the extension. Final submissions, purchases, and sign-ins are refused.
No WorkPods account is required. Provider keys stay in the extension; the MCP server returns results to the agent and does not receive those keys.
Links
Agent Ready by WorkPods · Source code · Privacy and permissions · CLI documentation · Apache-2.0 license
Built by WorkPods. Written by Shefali Munjal. See LICENSE and NOTICE for licensing and attribution.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.