AI coding assistants like Claude and Cursor are changing how developers interact with tools. Instead of switching between dashboards, terminals, and browser tabs, you describe what you need and the agent handles the rest. Website monitoring is a natural fit for this workflow. You should be able to ask "what changed on my competitor's pricing page this week?" and get an answer without opening a separate app.
PageCrawl is the only dedicated web monitoring tool with a purpose-built MCP (Model Context Protocol) server. This means Claude, Claude Code, Cursor, ChatGPT, OpenClaw, and any MCP-compatible AI tool can create monitors, check for changes, review diffs, and manage your monitoring setup through conversation. Creating monitors works on every plan, including Free.
This guide covers what the MCP server can do, practical examples, and how to set it up.
What Is MCP and Why Does It Matter for Monitoring
The Model Context Protocol is an open standard that lets AI assistants connect to external tools and data sources. Instead of the AI guessing about your monitoring data or asking you to copy-paste results, MCP gives it direct access to your actual monitors, change history, and configurations.
For web monitoring, this means:
- No context switching. Ask about changes without leaving your editor or terminal.
- Natural language queries. "Show me all price changes across my competitor monitors this month" works as a query.
- Automated workflows. AI agents can set up monitors, trigger checks, and process results as part of larger tasks.
- Living knowledge base. Your monitored pages become queryable data, not just an alert feed.
What You Can Do with the PageCrawl MCP Server
The MCP server exposes 13 tools that cover the full monitoring lifecycle. They split into three groups: read tools and the create tool work on every plan including Free, while a small set of on-demand action tools require a paid plan.
Reading and Searching (works on all plans, including Free)
- List workspaces and list monitors across all your workspaces, with optional search by URL, domain, or name
- Get monitor details including configuration, last check status, and tracking mode
- Get monitor history for any monitor, with date filtering
- Get latest values from tracked elements in a single batch of up to 50 monitors (current price, availability status, page content)
- Get the check diff for a specific check, showing exactly what changed
- Get changes since a given point in time, across your monitors
- List templates you can reuse when creating new monitors
Creating Monitors (works on all plans, including Free)
- Add a page monitor from a URL. The AI auto-detects the right tracking mode when you do not specify one, and supports 13 tracking modes (fullpage, content_only, reader, feed, price, specific_text, specific_number, seo, leaderboard, pdf_extract, http_status, json_path, ai_extract), page-interaction actions, tags, check frequency, and notification channels. New monitors default to screenshots on, cookie and overlay removal actions, and notifications auto-detected from your workspace's configured channels. Your plan's monitor-count limit still applies.
On-Demand Actions (require Standard plan or above)
- Trigger a check on any monitor without waiting for the schedule
- Manage tags by adding or removing them on a monitor (listing tags is free)
- Mark changes seen so detected changes are recorded as reviewed
- Update monitor defaults for a workspace (screenshots, notifications, frequency, AI focus)
On a Free account, the action tools return a clear error that links to the upgrade page, so nothing fails silently.
Tool reference
In conversation, the AI references these tools by their kebab-case names:
Read tools (all plans, including Free):
list-workspaceslists your workspaceslist-monitorslists monitors, optionally filtered by search termget-monitor-detailsreturns a monitor's configuration and statusget-monitor-historyreturns change history with date filteringget-latest-valuesreturns current tracked values for up to 50 monitors at onceget-check-diffreturns the formatted diff for a specific checkget-changes-sincereturns changes across monitors since a point in timelist-templateslists reusable monitor templates
Create tool (all plans, including Free):
add-page-monitorcreates a monitor from a URL with auto-detected or specified tracking mode
Action tools (Standard plan or above):
trigger-checkruns an immediate checkmanage-tagsadds or removes tags on a monitormark-changes-seenmarks detected changes as reviewedupdate-monitor-defaultsupdates a workspace's default settings
Practical Examples
Here are real prompts you can use with Claude, Claude Code, Cursor, ChatGPT, or any MCP client once the server is connected:
Creating Monitors (works on Free)
"Add a monitor for https://store.example.com/product-123
and track the price and availability"The AI calls add-page-monitor, auto-detects the price tracking mode, sets the elements, and applies the default settings. No need to configure selectors or choose between tracking types. This works on every plan, including Free, up to your monitor-count limit.
"Set up a monitor for our competitor's changelog at
https://competitor.com/changelog using reader mode,
check every 15 minutes, and notify the team Slack channel"One sentence produces a fully configured monitor with frequency, tracking mode, and notifications. The next time the competitor pushes an update, you get notified through your configured channels.
"Create monitors for each of these 5 competitor pricing pages"Batch creation in a single conversation turn. (Adding tags to them is a paid action, see below.)
Checking for Changes
"What changed across all my monitors in the last 7 days?"Returns a summary of recent changes with priorities, timestamps, and affected monitors.
"Show me the diff for the last change on the Acme Corp pricing page"Returns a formatted text diff showing exactly what was added, removed, or modified.
Research and Analysis
"What is the current price on each of my competitor product monitors?"Pulls the latest tracked values across all price monitors, giving you a competitive pricing snapshot without visiting any dashboards.
"Have any of my compliance monitors changed in the last 30 days?
Show me the details."Filters by tag or search term, then retrieves change history and diffs for matching monitors.
On-Demand Actions (paid plans)
"Trigger a check on all my competitor pricing monitors right now"Calls trigger-check so you get fresh results within minutes instead of waiting for the schedule. Handy before a meeting or review.
"Tag the 5 monitors I just created as 'q2-pricing'"Uses manage-tags to organize monitors after creation, so later queries can filter by tag.
"Mark all the changes on my newsletter monitors as seen"Calls mark-changes-seen to clear reviewed changes from your queue, keeping your unseen list focused on what you have not looked at yet.
How It Compares to Other Approaches
Several scraping tools offer MCP servers (Firecrawl, Apify, Bright Data), but these are built for one-time data extraction, not ongoing monitoring. They can fetch a page right now, but they cannot:
- Track changes over time. No history, no diffs, no "what changed since last week."
- Send notifications. No email, Slack, Discord, or Teams alerts when something changes.
- Score change importance. No AI priority scoring to filter noise from signal.
- Handle protected sites. No managed proxy infrastructure or bot protection handling.
A scraping MCP server answers "what does this page say right now?" PageCrawl's MCP server answers "what has changed, when, and does it matter?"
No other dedicated monitoring tool (Visualping, Distill.io, Hexowatch, ChangeTower, ChangeDetection.io) offers MCP integration. If you want AI agents to interact with your web monitoring data, PageCrawl is currently the only option.
Setting Up the MCP Server
The PageCrawl MCP server is hosted at https://pagecrawl.io/mcp and is available on all plans, including Free. Free accounts can create monitors and use every read tool, with your plan's monitor limit still applying. Paid plans (Standard or above) add the action tools: triggering on-demand checks, managing tags, marking changes seen, and updating monitor defaults.
Connect with OAuth (recommended, no token)
The fastest way to connect is OAuth 2.0, which works on Free accounts:
Step 1: Sign up or log in at pagecrawl.io.
Step 2: In your AI tool's MCP or connector settings, add a new server with the URL https://pagecrawl.io/mcp, then sign in once when prompted. There is no API key or token to paste, and nothing to install locally.
Step 3: Verify the connection by asking your assistant to "list my PageCrawl monitors."
Note: Free accounts are rate-limited to 30 requests per minute, which is plenty for interactive use.
For Claude and Claude Code
In Claude.ai, add PageCrawl under connectors and sign in with OAuth. Once connected, all 13 monitoring tools are available in your conversations. Claude Code (and other headless or CLI clients) can use the same OAuth flow, or alternatively authenticate with a personal API token from Settings > API, which is convenient for scripts and automation.
For Cursor
Connect the PageCrawl MCP server through Cursor's MCP configuration using the https://pagecrawl.io/mcp URL. The monitoring tools integrate naturally into your coding workflow, so you can check for API changes or competitor updates while working on code. Like Claude Code, Cursor can also use a personal API token from Settings > API.
For ChatGPT, OpenClaw, and other MCP clients
ChatGPT, OpenClaw, and any other MCP-compatible tool connect the same way: point them at https://pagecrawl.io/mcp and authenticate with OAuth, or use a personal API token for headless setups. The same 13 tools are exposed regardless of client.
Tips for Getting the Most Out of It
Start with what you already monitor. If you have existing monitors, use the MCP server to query them: "summarize all changes from last week" or "which monitors have unseen changes?" This is the fastest way to see the value.
Use tags for organization. Tag monitors by project, client, or category. Then query by tag: "show changes on all monitors tagged 'client-acme'."
Combine with AI importance filtering. Set a "What matters" prompt on your monitors (e.g., "Focus on price changes, ignore navigation updates"). The AI will score changes accordingly, and when you query via MCP, the priority scores help you focus on what matters.
Trigger checks on demand (paid). Before a meeting or review, trigger checks on key monitors: "trigger a check on all my competitor pricing monitors." Results are available within minutes. On-demand checks are a paid action, but creating and reading monitors is free.
Choosing your PageCrawl plan
PageCrawl's Free plan lets you monitor 6 pages with 220 checks per month, which is enough to validate the approach on your most critical pages. Most teams graduate to a paid plan once they see the value.
| Plan | Price | Pages | Checks / month | Frequency |
|---|---|---|---|---|
| Free | $0 | 6 | 220 | every 60 min |
| Standard | $8/mo or $80/yr | 100 | 15,000 | every 15 min |
| Enterprise | $30/mo or $300/yr | 500 | 100,000 | every 5 min |
| Ultimate | $99/mo or $999/yr | 1,000 | 100,000 | every 2 min |
Annual billing saves two months across every paid tier. Enterprise and Ultimate scale up to 100x if you need thousands of pages or multi-team access.
At an engineering hourly rate, Standard at $80/year pays for itself the first time you catch a breaking API change, a deprecated endpoint, or a silent config change before it takes down production. 100 monitored pages is enough to cover the changelogs and docs of every third-party API your stack depends on. Enterprise at $300/year adds higher check frequency, 500 pages, and full API access. All plans include the PageCrawl MCP Server, which plugs directly into Claude, Cursor, and other MCP-compatible tools. Developers can ask "what changed in the Stripe API docs this month?" and get a summary pulled from your own monitoring history. AI assistants can create monitors through conversation on every plan, including Free, and paid plans add on-demand checks and monitor management, turning your tracked pages into a living knowledge base instead of a pile of alert emails.
Getting Started
Start with 3-5 pages you already check manually: a competitor's pricing page, an API changelog, a regulatory page, or a product you are waiting to restock. Set them up through the MCP server in a single conversation. Run it for two weeks and see how many manual checks it replaces.
Once you see the value, expand to cover your full competitive landscape, compliance requirements, or technical dependencies. The MCP server makes scaling from 5 monitors to 100 as simple as describing what you want to track.
PageCrawl's free tier includes 6 monitors and MCP access. Create your account and connect the MCP server to start monitoring from your AI assistant.