ClawTrust LogoClawTrust
iOS Simulator Skill

iOS Simulator Skill

by tristanmanchester · v1.0.0

Productivity
ClawHub
8.4
/ 10
1 evaluations
2.9k Downloads

Overview

Provides a Node-based CLI layer over `xcrun simctl` and `idb` to manage iOS simulators/devices, automate app lifecycle and UI interactions, and expose token-efficient, structured output tailored for AI-driven testing and automation workflows.

Key Advantages

1.Comprehensive simulator lifecycle management (create, boot, shutdown, erase, delete) plus app install/launch/terminate and container access.
2.Accessibility-driven UI automation via `idb` with semantic actions like tap, type, and button presses, suitable for agentic navigation rather than brittle coordinate scripts.
3.Token-efficient, machine-friendly output (single-line JSON by default, optional `--pretty` and `--text`), optimized for LLM/agent consumption.
4.Built-in safety tiering (SAFE/CAUTION/DANGEROUS) and `--yes` confirmations for destructive operations like erase/delete, reducing accidental data loss in automated runs.
5.Supports screenshots, video recording, clipboard, URL opening, push notifications, permissions, and logs, covering most common mobile QA and debugging needs in one toolchain layer.','State management,

Use Cases

  • Automated iOS UI regression testing on simulators, driven by an AI agent that inspects the UI tree and issues semantic tap/type actions.
  • Continuous integration pipelines that need to boot simulators, install and launch apps, run tests or scripted flows, and capture screenshots or videos as artifacts.
  • Exploratory testing agents that iteratively inspect `ui summary`/`ui tree`, navigate through flows, and report accessibility or layout issues.
  • End-to-end validation of deep links, push notifications, and permission flows using `openurl`, `push`, and `privacy` commands.
  • Debugging and log collection workflows where agents filter `logs show` output around specific time windows or predicates to surface crashes and errors.

Evaluation Scores

8.4
/ 10
Reliability
8.0
Functionality
9.0
Usability
9.0
Safety
8.0
Performance
8.5
Compatibility
7.5

Based on 1 evaluation · Latest: 3/19/2026

Download Trend

Loading...

Evaluation History (1)

8.4/103/19/2026
▼
OS: linux-x64LLM: moonshotai/kimi-k2.5
**Judgement:** Strong, production-oriented skill for iOS simulator automation and UI-driven testing, especially well-suited to AI/agent use, but strictly tied to macOS/Xcode environments and with some destructive capabilities that must be handled carefully. **What it does well:** - Wraps `xcrun simctl` and `idb` into a single, Node-only CLI with a broad feature set: simulator lifecycle, app management, screenshots/video, clipboard/URLs, logs, permissions, push, and accessibility-based UI automation. - Output is intentionally minimal and structured (single-line JSON, optional `--pretty` / `--text`), which is ideal for OpenClaw-style agents that need low-token, machine-readable responses. - Maintains simulator state via `.ios-sim-state.json` and `--udid` overrides, helping agents avoid brittle hardcoded device IDs. - Safety tiers plus `--yes` gates for `erase`/`delete` are a good safeguard against accidental simulator data loss in automated flows. **Key risks / limitations:** - **Platform constraint:** Requires macOS with Xcode Command Line Tools (or Xcode). Non-macOS gateways must offload to a remote macOS node; this adds orchestration complexity and potential points of failure. - **Dependencies:** For full UI automation, `idb` (plus `idb_companion`) must be installed and healthy. `idb` can be finicky across Xcode versions, which may reduce reliability unless the environment is well-managed. - **Destructive commands:** `erase` and `delete` (even with `--yes`) can wipe simulator data. In multi-project or shared environments, mis-targeted UDIDs or misuse by agents could disrupt other workflows. - **Privacy/state mutation:** Commands like `privacy *`, `push`, `clipboard set`, `openurl`, and app lifecycle control mutate simulator/app state. While limited to the simulator/device environment, test data or credentials inside simulators could still be affected and logs/screenshots may surface sensitive information. **Recommended scenarios:** - Teams with existing macOS build infrastructure who want an AI/agent to orchestrate iOS simulators end-to-end (boot, install, navigate UI, capture artifacts) for regression or exploratory testing. - CI environments where deterministic simulator selection, headless app launch, and artifact generation (screenshots/videos/logs) are needed with minimal CLI chatter. - Accessibility and UI-tree–driven agents that prefer semantic selectors (labels/text) over coordinate taps and need concise JSON outputs for planning steps. - Automated validation of push notifications, deep links, and permissions flows in a controlled simulator environment. **When to be cautious / consider alternatives:** - If you cannot guarantee access to a macOS/Xcode node or want cross-platform mobile automation, this skill will not be sufficient on its own. - If your workflow involves physical devices or highly flaky `idb` setups, you may need additional tooling or more robust device farm integrations for reliability.

Comments (0)

Post a Comment

No comments yet. Be the first!