Skip to main content
Glama

Create monitor

create_monitor

Create a website monitor that keeps checking a page on a schedule after the conversation ends, captures screenshots and HTML, and alerts only when meaningful changes occur.

Instructions

Create a new website monitor that keeps checking after this conversation ends: monity.ai checks the page on a schedule in its own cloud with a stealth browser (Camoufox, an anti-detection Firefox) and smart proxies that switch to residential routes when a site blocks the check, keeps a full-page screenshot and a full HTML copy of the page for every check, and alerts only when the change matters (requires a write-scope API key; subject to the plan's monitor limit). compareType AUTO (default) covers most pages; VISUAL for pixel comparison, TEXT for keyword monitoring, HTML for markup diffs. Express the full intent in one call: use notifyWhen for a natural-language alert condition ('price drops below $500'), or notifyConditions + outputExtractor/jsonPath for deterministic threshold alerts without AI.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "websiteUrl": "<websiteUrl>"
    },
    "name": "create_monitor"
  }
}

Input Schema

NameRequiredDescriptionDefault
nameNoMonitor name (defaults to the URL host/path)
labelsNoLabel names to file the monitor under (up to 10); existing labels with the same name are reused.
skillsNoExtra checks to run on the page at every check, each alerting on its own rules, on top of change monitoring: skill names from list_skills ('perf', 'seo', 'a11y', 'links', 'xss', 'tls', 'scripts', 'privacy', …) or plain words ('core web vitals', 'ssl', 'gdpr'). Add one only when the user asks for what it does. A skill the plan does not include is not added and comes back in skillsSkipped.
actionsNoOrdered list of interactions to perform before each check, e.g. [{"type":"CLICK","selector":"#accept-cookies"},{"type":"WAIT","value":1500},{"type":"CLICK","selector":".load-more"}]. Use this when the content to monitor is behind a cookie banner, a tab, a "show more" control, or a dropdown. Max 20 steps. Script and JS/CSS injection steps are dashboard-only and rejected here.
jsonPathNoFor JSON API endpoints: JSONPath to the value to monitor, e.g. $.data.items[0].price. Starts with '$'.
keywordsNoTEXT monitors only: comma-separated keywords to look for, e.g. "sold out, unavailable"
selectorNoCSS selector of ONE element to monitor instead of the whole page (e.g. '.product-price' or '#stock-status'). Makes an element monitor that is found by this selector on every check; leave compareType unset.
frequencyNoMinutes between checks, as a string: '5', '15', '30', '60' (hourly), '180', '360', '720', '1440' (daily), '10080' (weekly), '302400' (monthly). Omit for the plan default. A frequency faster than the plan allows is raised to the plan minimum.
notifyWhenNoNatural-language alert condition evaluated by AI on every check, e.g. 'notify me when the price drops below $500' or 'alert when the Pro plan changes'. Mutually exclusive with notifyConditions.
websiteUrlYesAbsolute http(s) URL of the page to monitor
compareTypeNoComparison mode (default AUTO)
skillAlertsNoAlert rules per skill, only when the user names a threshold or a rule, e.g. {"perf": {"lcp": 3}}: a number sets the rule's threshold and turns it on (seconds for timings, MB for page weight, days for expiry, 0-100 for scores), true/false turns a rule on or off, null resets it. Rule ids are in list_skills. A skill named here is added to the monitor.
workspaceIdNoWorkspace to create the monitor in (from list_workspaces). Defaults to the key owner's personal workspace, which on a business account is NOT visible to the rest of the team — pass this to put it somewhere shared.
notifyEmailsNoExtra email addresses to CC on alerts (workspace members are emailed by default)
keywordsActionNoTEXT monitors only: alert when keywords are ADDED to the page (default) or REMOVED from it
outputExtractorNoHow to extract the monitored value from the page text: NUMBER pulls the first number (good for prices), REGEX applies outputExtractorPattern, NONE keeps the full text.
notifyConditionsNoDeterministic alert conditions on the extracted output (no AI credits). All conditions must pass. Requires the output to change. Mutually exclusive with notifyWhen. Not available for HTML monitors.
notifyWebhookUrlNoWebhook URL to POST alerts to (HMAC-signed with the account webhook secret)
outputExtractorPatternNoRegular expression for outputExtractor: REGEX — bare pattern or /pattern/flags; the first capture group (or full match) becomes the monitored output, e.g. 'In stock: (\d+)'

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.5.0

TDQS

A4.1/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden and does so: write-scope API key required, plan monitor-limit enforcement, checks run in the vendor's cloud after the conversation ends (not in the agent's session), artifacts retained per check, alerts fire only on meaningful change, out-of-plan skills silently skipped and returned in skillsSkipped. These are non-obvious side effects an agent could not get from the schema.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness3/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The opening clause is front-loaded and informative, but the block re-states schema facts (compareType AUTO default, per-mode meanings that largely duplicate the enum description) and piles many ideas into one dense run-on. It is informative rather than bloated, yet some sentences do not earn their place against the schema.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 19-parameter, nested-object tool with no output schema, the description covers the behaviors that matter (auth scope, plan limits, skill skipping) and most parameter interactions. It never says what a successful call returns (e.g. the created monitor's id for later get_monitor/update_monitor calls), which is the main remaining gap.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3; the description still adds cross-parameter meaning the schema does not, e.g. pairing notifyConditions with outputExtractor/jsonPath for deterministic thresholds, and the VISUAL/TEXT/HTML choice in terms of the monitoring goal rather than the enum values.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

"Create a new website monitor" gives a specific verb and resource, and the following clauses explain what the monitor does (scheduled cloud checks with stealth browser/proxies, screenshots and HTML snapshots, change-only alerts). It does not name a sibling such as update_monitor or create_public_tracker, so an agent must infer the boundary from the verb alone, but the purpose itself is unambiguous.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives real decision guidance: "Express the full intent in one call", "compareType AUTO (default) covers most pages", and the notifyWhen-vs-notifyConditions split for AI vs deterministic alerts. It does not say when to prefer this over update_monitor or create_public_tracker, so the exclusions/alternatives half is missing.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.