Skip to main content
Glama

Robot Actions — Remote Device Control

ios_safari_evaluate

Evaluate a JavaScript expression in the context of a Safari page on an iOS device via the Web Inspector. Uses Runtime.evaluate internally. Returns the result as a string (JSON-stringified for objects), followed by a second block naming the tab it actually ran on — check that if a result looks like it came from the wrong page. Omit pageId to auto-pick the frontmost Safari tab. WAITING: UI frameworks paint a frame or two AFTER the click that triggers them, so an evaluate fired immediately can read the DOM before the menu/overlay/row exists. Use waitForSelector to block until an element appears (and waitMs to let it settle) INSTEAD of splitting the work into a separate poll call — both wait inside this same call on the same tab. DO NOT CLICK WITH THIS TOOL. A click dispatched from JavaScript — .click(), or dispatchEvent of a MouseEvent — is an untrusted event (isTrusted false). Component frameworks and anything gated on a real user gesture (file pickers, clipboard, autoplay, anti-bot checks) may ignore it, so the evaluate SUCCEEDS and the page does nothing — a failure that looks like a passing call. Use webpage_click instead: it has the device tap its own screen at the element, delivering a real trusted touch (pointerdown, touchstart, mousedown, click). Same for typing — webpage_type.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
udidYesiOS device UDID
pageIdNoCDP target id from ios_safari_list_pages. Omit it to use the page ios_safari_navigate last landed on (or, failing that, the first loaded page) — only pass it to target a DIFFERENT tab
waitMsNoSettle delay applied just before evaluating (and AFTER waitForSelector matches, when both are given). Use for animations that have started but not finished. Maximum 10000 ms
expressionYesJavaScript expression to evaluate
waitTimeoutMsNoHow long waitForSelector may poll. Default 5000 ms, maximum 15000 ms
platformVersionNoIgnored (kept for compatibility)
waitForSelectorNoCSS selector to wait for BEFORE evaluating. Polls until it matches, then runs the expression. Errors if it never appears within waitTimeoutMs — it will not evaluate anyway and hand back a null that reads like a real answer

TDQS

A4.7/5.0
Behavior5/5

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

The description discloses return format (string plus second block naming the actual tab), warns that results may come from the wrong page, explains frontmost tab auto-pick behavior, and details the untrusted event issue (isTrusted false) meaning clicks may silently do nothing. It also notes waitForSelector's error behavior returning a null that looks like a real answer. With no annotations, this transparency is particularly valuable.

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?

The description is long but every segment earns its place, moving from core purpose to return format to critical behavioral warnings, with clear section markers like 'WAITING:' and 'DO NOT CLICK WITH THIS TOOL.' It's front-loaded and structured for quick scanning.

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 complex tool with no output schema, the description explains return value structure, tab selection behavior, waiting semantics, and dangerous misuse patterns. It covers the main edge cases (wrong tab, untrusted events, null on timeout) and integrates with sibling tool alternatives, making it effectively complete.

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

Parameters3/5

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 adds contextual usage guidance for waitForSelector and waitMs but doesn't introduce new parameter semantics beyond what the schema already provides. The pageId reference is simplified and slightly inconsistent with the schema's 'last navigated' rule, but not a significant addition.

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 clearly states that the tool evaluates a JavaScript expression in a Safari page on an iOS device via the Web Inspector, naming the specific verb and resource. It distinguishes itself from sibling evaluate tools (android_devtools_evaluate, web_evaluate) by tying to iOS Safari, and adds scope by mentioning return format and pageId behavior.

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 explicitly instructs not to click with this tool and names webpage_click as the alternative for trusted events, and webpage_type for typing. It also advises using waitForSelector within the same call instead of a separate poll call, giving clear when-to-use and when-not-to-use guidance.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

TDQS

B3.1/5.0
Disambiguation2/5

The set contains near-identical duplicate families: web_* and playwright_* expose ~15 pairs of the same desktop-grid-browser operations (web_get_text/playwright_get_text, web_reload/playwright_reload), and screenshot/log/network/mock capabilities each have 5-8 entry points (device_screenshot vs android_mjpeg_screenshot vs ios_screenshot vs ios_fast_screenshot vs web_screenshot vs webpage_screenshot vs session_screenshot). Many individual descriptions carefully draw boundaries (devtools vs traffic, HID vs session), but an agent cannot reliably distinguish web_* from playwright_*, and ios_screenshot/ios_fast_screenshot/ios_mjpeg_screenshot blur together.

Naming Consistency2/5

The prefix scheme is broken: Android functionality is split arbitrarily between android_* and device_* (device_screenshot vs android_mjpeg_screenshot), the desktop browser gets two parallel prefixes (web_* and playwright_*), and verbs vary across equivalents (device_navigate_url vs web_navigate vs ios_safari_navigate). session_* uses bare verbs (session_url, session_back), and the same concept gets different names (ios_clipboard_get_hid vs ios_get_pasteboard; device_screen vs ios_orientation).

Tool Count1/5

333 tools is an extreme count by any measure — far beyond the 50+ threshold — and much of the bulk is duplicative (the web_*/playwright_* pairs alone double ~15 slots) or out-of-scope for a device-control server (TestRail, Jira, AzDO, agent memory, secret variables, feedback). Even granting that remote device control + test automation is a broad domain, this surface will devastate agent context budgets and is impossible to navigate coherently.

Completeness4/5

The core device-control and test-automation domain is remarkably thorough: Android and iOS each have full interaction, app-lifecycle, file, network/proxy, performance, crash, accessibility, recording, and replay coverage, with CRUD lifecycles for flows, suites, app uploads, TestRail cases, and visual-review baselines. Minor gaps exist at the margins — Jira/AzDO lack update/transition/comment operations, and iOS cannot open/close tabs — but the central workflows have no dead ends.

Resources