Robot Actions — Remote Device Control

iOS drag drop

ios_drag_drop

Drag-and-drop on an iOS device: hold at (x1,y1) for holdMs to grab, then move to (x2,y2) over durationMs. Distinct from ios_swipe (no explicit hold). Use for home-screen icon reorder, drag-into-folder, slide-to-confirm. Coordinates are screen points. Works without an automation session when the direct input path is available; falls back to one otherwise (see input).

Input Schema

TableJSON Schema

Name	Required	Description
`x1`	Yes	Start X (screen points)
`x2`	Yes	End X (screen points)
`y1`	Yes	Start Y (screen points)
`y2`	Yes	End Y (screen points)
`norm`	No	If true, x/y (and path points) are 0..1 FRACTIONS of the screen — pass `pixelInScreenshot ÷ screenshotSize` directly (scale-free, matches the frontend, avoids point-guessing). Default false = iOS screen points.
`udid`	Yes	iOS device UDID
`input`	No	Input path. 'auto' (default) tries the direct input path first — the one that needs no automation session — and falls back to the session only when the direct path is not usable right now (the response says which one ran). On a device set to direct-input-only, an unusable direct path is reported as an error instead of silently switching to the session. 'hid' forces the direct path with no fallback, reporting 'no HID available' on iOS 17.x or with no tunnel. 'session' forces the automation session even when the direct path would have worked.
`holdMs`	No	Hold-in-place duration before motion in ms (default: 500)
`durationMs`	No	Motion duration from start to end in ms (default: 400)

Schema Changelog

Changes observed during successful MCP inspections.

Changed2 schema fields changedSep 28, 2026

addedInput schema / properties / input

Added value: +{
+  "description": "Input path. 'auto' (default) tries the direct input path first — the one that needs no automation session — and falls back to the session only when the direct path is not usable right now (the response says which one ran). On a device set to direct-input-only, an unusable direct path is reported as an error instead of silently switching to the session. 'hid' forces the direct path with no fallback, reporting 'no HID available' on iOS 17.x or with no tunnel. 'session' forces the automation session even when the direct path would have worked.",
+  "enum": [
+    "auto",
+    "hid",
+    "session"
+  ],
+  "type": "string"
+}

addedInput schema / properties / norm

Added value: +{
+  "description": "If true, x/y (and path points) are 0..1 FRACTIONS of the screen — pass `pixelInScreenshot ÷ screenshotSize` directly (scale-free, matches the frontend, avoids point-guessing). Default false = iOS screen points.",
+  "type": "boolean"
+}

First observedAug 4, 2026

TDQS

A4.7/5.0

Behavior5/5

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

Beyond the minimal safety annotations (readOnlyHint=false, destructiveHint=false), the description discloses non-obvious runtime behavior: the 'auto' path tries direct input first and falls back to a session, direct-input-only devices report an error rather than silently switching, and the response indicates which path ran. This is exactly the kind of behavioral context annotations cannot carry.

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?

Roughly five sentences, front-loaded with the action mechanics before alternatives and input-path details. It is dense and each sentence carries information, though the input-path sentence is long and the density could be trimmed slightly.

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 9-parameter gesture tool with no output schema and full schema coverage, the description covers mechanics, coordinate semantics, use cases, sibling differentiation, and fallback behavior. It stops short of describing failure modes (e.g., no element at coordinates) or the response shape beyond noting it reports which input path ran.

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, but the description adds meaning by assigning temporal roles to the parameters — x1/y1 are the grab point held for holdMs, x2/y2 are the target reached over durationMs — which the schema does not express. It also reinforces that coordinates are screen points, though that partially repeats the schema.

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?

The description states a specific verb and resource ('Drag-and-drop on an iOS device') and describes the exact mechanics: hold at (x1,y1) for holdMs to grab, then move to (x2,y2) over durationMs. It explicitly distinguishes itself from the ios_swipe sibling ('Distinct from ios_swipe (no explicit hold)'), so an agent can route between them without opening schemas.

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?

It names concrete use cases ('home-screen icon reorder, drag-into-folder, slide-to-confirm') and explicitly contrasts with the nearest alternative, ios_swipe, giving the condition that selects this tool (explicit hold required). The input-path guidance further clarifies when the direct path vs session path applies.

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

Give agents eyes on the webHosted screenshot MCP. Send a URL, get a pixel-perfect image. $0.002 per grab. Works with Claude, Cursor, and Codex.grabbit.liveAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?