Robot Actions — Remote Device Control

webpage_screenshot

Screenshot the web page in the device browser (iOS Safari or Android Chrome, auto-detected). ANDROID: a clean capture of the web CONTENT — pass selector or ref to capture just that element (scrolled into view). iOS: element-clipped web capture is unavailable, so this returns the full DEVICE frame (includes the Safari chrome/status bar); selector/ref clipping is Android-only and is ignored on iOS (a note is returned). Returns a JPEG image.

Input Schema

TableJSON Schema

Name	Required	Description
`ref`	No	Android only: capture just this element, named by a ref from a page-elements listing
`udid`	Yes	Device UDID / serial (iOS or Android)
`pageId`	No	Target page/tab id — auto-picked when omitted
`socket`	No	Android only: abstract unix socket name (default: chrome_devtools_remote)
`selector`	No	Android only: capture just this element (CSS selector), scrolled into view
`platformVersion`	No	Ignored (kept for compatibility)

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedAug 9, 2026

addedInput schema / properties / ref

Added value: +{
+  "description": "Android only: capture just this element, named by a ref from a page-elements listing",
+  "type": "string"
+}

First observedAug 4, 2026

TDQS

A4.7/5.0

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 succeeds. It discloses auto-detection of the browser, Android-only element clipping, iOS returning the full device frame including Safari chrome/status bar, selector/ref being ignored on iOS with a note, and the JPEG output format. These are behavioral traits an agent must know to set expectations correctly, and they are all explicitly stated.

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

Conciseness5/5

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

Two high-density sentences elegantly cover purpose, platform differences, limitations, and return format. The purpose is front-loaded, and every clause carries distinct information without repetition. No fluff or redundant restatement of the tool name.

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?

The tool's complexity lies in its platform-specific behavior, and the description captures all of it: Android vs. iOS output, selector/ref limitations, the note returned, and auto-detection. All six parameters are schema-documented, the return type is stated, and the description is complete enough for an agent to invoke it correctly without additional probing.

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?

The input schema already covers 100% of parameters with descriptions, giving a baseline of 3. The description adds critical runtime semantics: selector and ref are Android-only and are ignored (with a note) on iOS. It also clarifies auto-detection, the automatic picking of pageId, and that platformVersion is ignored, which goes beyond the schema's static param docs and helps an agent predict actual behavior.

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: 'Screenshot the web page in the device browser.' It is clearly differentiated from generic screenshot tools by specifying the mobile browser context (iOS Safari or Android Chrome) and the return type (JPEG). The description also distinguishes its behavior from sibling screenshot tools by explaining the difference between Android content capture and iOS device-frame capture.

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?

Provides clear platform-specific usage instructions: on Android, use selector/ref for element capture; on iOS, element clipping is unavailable, so the full device frame is returned, and a note is returned if selector/ref are supplied. The description tells the agent exactly what will happen on each platform, but does not explicitly name sibling alternatives (e.g., device_screenshot or web_screenshot) or state when to prefer them, which leaves some routing responsibility to the agent.

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

All Robot Actions — Remote Device Control tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?