Cactal

Capture an external page

webResearch_screenshot

Read-onlyIdempotent

Renders an external public page in a real browser and returns the image, to study a reference site or the user's existing site: its layout, hierarchy, density, and feel. Consent banners and overlays are dismissed first. Focus with selector (plus padding) or clip and page down with scrollTo; external pages render in their own color scheme and never full page. Capture your own website with websitePreview.capture. Charged in AI credits to the website's organization.

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": {
      "websiteId": "<websiteId>",
      "url": "<url>"
    },
    "name": "webResearch_screenshot"
  }
}

Input Schema

Name	Required	Description	Default
`url`	Yes	Absolute public http(s) URL of the page to capture.
`clip`	No	Crop to this rectangle, in CSS pixels from the top-left of the rendered page at the chosen viewport; it is scrolled into view and must fit within the viewport. Only for regions no selector frames.
`delayMs`	No	Extra settle time before capturing, in milliseconds.
`padding`	No	CSS pixels of surrounding page to include on every side of the `selector` match.
`scrollTo`	No	Scroll to a CSS selector or an absolute Y offset before capturing.
`selector`	No	Crop to the first element matching this CSS selector, scrolled into view. Preferred: it follows the layout at every viewport.
`viewport`	No	Device class to render at.	desktop
`websiteId`	Yes	A website you can edit; its organization pays for the call in AI credits.
`waitForSelector`	No	Wait for this CSS selector to match before capturing.
`freezeAnimations`	No	Fast-forward CSS animations and transitions to their end state so captures are deterministic.

Schema Changelog

Changes observed during successful MCP inspections.

First observedNov 2, 2026

TDQS

A4.7/5.0

Behavior4/5

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

Annotations already cover safety (readOnly, idempotent, openWorld, non-destructive), and the description adds real context: consent banners/overlays are dismissed automatically, external pages render in their own color scheme, full-page capture is never produced, and the call is charged in AI credits to the website's organization. The only gap is nothing about failure modes for pages that don't load.

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

Conciseness4/5

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

Purpose and consent/overlay behavior are front-loaded in the first two sentences, with framing and cross-tool routing after. Dense but nearly every clause carries information; the credit-charging note is arguably better placed in annotations or the `websiteId` description where it already appears.

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

Completeness5/5

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

For a 10-parameter, nested-object tool with no output schema, the description covers what is returned (an image), the viewport/framing strategy, a hard capability limit (never full page), and cost attribution. All parameters are described in the schema, so nothing an agent needs to call it is missing.

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

Parameters5/5

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 exceeds that by explaining the relationship between parameters — `selector` is preferred and should be paired with `padding`, `clip` is a fallback for regions no selector frames, and `scrollTo` pages down — not just restating names. That relational guidance is unavailable from the schema alone.

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

Purpose5/5

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

States a specific verb and resource ('renders an external public page in a real browser and returns the image') plus the intent (study layout, hierarchy, density, feel). It also explicitly separates itself from the sibling `websitePreview.capture`, which captures your own site.

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

Usage Guidelines5/5

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

Names the alternative tool and the condition that selects it ('Capture your own website with websitePreview.capture'), and the word 'external public' defines what is out of scope. It also gives concrete framing guidance — focus with `selector` (+`padding`) or use `clip` only for regions no selector frames, and `scrollTo` to page down.

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

All Cactal tools Inspect

Try in Browser

Claude Code, inside the Godot 4 editorFlockbay puts an AI agent in the real Godot 4 editor: it writes GDScript, runs your game and fixes what breaks. Free, or bring your own Claude Code or Codex.flockbay.comAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?