← Back to Skills Marketplace
Mobilerun
by
johnmalek312
· GitHub ↗
· v2.1.0
· MIT-0
796
Downloads
3
Stars
3
Active Installs
22
Versions
Install in OpenClaw
/install mobilerun
Description
Load this skill whenever the user wants to control, automate, or interact with a phone or mobile device. This includes: tapping, swiping, typing, taking scre...
Usage Guidance
This skill appears internally consistent with its purpose, but it grants an agent the ability to view and control a user's Android device (screenshots and UI trees may contain sensitive personal data). Before installing: 1) Confirm you trust the Mobilerun service and the source of the skill (there's no homepage listed here); verify the Portal APK and GitHub repos referenced in the docs. 2) Only provide the MOBILERUN_API_KEY via your host platform (OpenClaw) — do not paste it in chat. 3) Be cautious with webhooks/targetUrls — subscribing to an external webhook could send device/task data to a third party. 4) If you plan to use this for highly sensitive apps (banking, private messages), consider whether you want automated agent actions to access that data and ensure you can revoke keys quickly. If you want deeper assurance, ask the publisher for an authoritative homepage, privacy/security contact, and a reproducible GitHub release URL for the Portal APK.
Capability Analysis
Type: OpenClaw Skill
Name: mobilerun
Version: 2.1.0
The 'mobilerun' skill bundle is a comprehensive and well-documented interface for a mobile automation platform. It provides the AI agent with tools to control Android devices via the Mobilerun API (api.mobilerun.ai), including screen observation, UI interaction, and autonomous task execution. The instructions in SKILL.md and references/security.md explicitly prioritize user privacy and credential safety, instructing the agent never to reveal the API key or share sensitive screen data with third parties. While the skill requires high-risk permissions (Accessibility Service on the target device) and handles sensitive data (screenshots), these are essential for its stated purpose, and the documentation provides transparent setup and security guidelines.
Capability Tags
Capability Assessment
Purpose & Capability
Name/description (phone control, tapping/swiping, screenshots, UI tree, tasks) align with the declared dependency on a single Mobilerun API key and curl/jq. Requested binaries and the single env var (MOBILERUN_API_KEY) are appropriate for a curl-driven REST integration with a hosted mobile-control service.
Instruction Scope
SKILL.md instructs the agent to call the Mobilerun API (GET/POST/DELETE) and to fetch screenshots and UI trees — all expected for remote device control. These operations can surface highly sensitive user data (messages, banking screens). The skill explicitly warns not to reveal the API key and to avoid exposing internal tokens and screen data to others; that is appropriate, but operators should be aware screenshots/UI trees are accessible and must be handled carefully.
Install Mechanism
Instruction-only skill with no install spec or downloaded code; lowest-risk delivery. It relies on curl and jq being present, which is consistent with the provided curl examples.
Credentials
Only one environment variable (MOBILERUN_API_KEY) is required and is justified by the described API usage. No unrelated credentials or system config paths are requested.
Persistence & Privilege
always is false, the skill does not request permanent system presence, and it does not modify other skills or system-wide configuration. Autonomous model invocation is allowed by default (normal) but not elevated here.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install mobilerun - After installation, invoke the skill by name or use
/mobilerun - Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.1.0
Restructured into references/ subfolder (troubleshooting, use-cases, security, setup-and-billing, changelog). Added searchable tags, Unicode keyboard support docs, iOS coming soon note, task queuing improvements, social proof and resource links.
v2.0.1
Mobilerun v2.0.1 brings expanded guidance for setup and device management.
- Clarifies when and how to check for connected, ready devices before taking action.
- Offers precise user instructions if no device is ready, including Portal setup steps.
- Details best practices for privacy and API key use—never expose keys to the user.
- Adds information on cleaning up cloud devices to avoid unnecessary credit use.
- Improves context gathering recommendations for more autonomous task execution.
- Enhances clarity on user-relevant device info and error reporting.
v2.0.0
Mobilerun 2.0.0
- Added jq as a required binary dependency for skill operation.
- Enhanced privacy guidelines: clarified that screenshots and UI tree may reveal sensitive data, and added instructions never to share, log, or print device content or API keys.
- Improved documentation on cloud device cleanup: emphasized that deleting a device immediately interrupts all associated tasks and they must be allowed to finish first.
- Updated instructions for submitting multiple tasks to a temporary cloud device; clarified task queuing and device lifecycle.
- Minor restructuring and improved clarity in workflow and usage instructions.
v1.0.18
Declare curl as a required binary in metadata; add input handling section to security.md
v1.0.17
Add input sanitization guidance for API path parameters; expand security.md with input handling section
v1.0.16
Move security and privacy documentation to dedicated security.md; lighten inline warnings in SKILL.md and reference.md
v1.0.15
Updated skill content and documentation
v1.0.14
Remove agent-side API key persistence, let OpenClaw handle credentials. Drop requires.config and the resolve-the-key instruction — OpenClaw injects MOBILERUN_API_KEY via primaryEnv before the skill loads.
v1.0.13
- Documentation clarified: when listing device info, only internal fields such as streamUrl, streamToken, etc. must remain hidden unless the user asks for technical details.
- Removed mention of an unavailable "socket toggle" and other undocumented controls.
- Overall user guidance made stricter: only show actionable information for device connectivity and status.
v1.0.12
- Updated documentation in api.md and subscription.md.
- No functional changes to the API or user interface.
- Documentation improvements clarify platform API and subscription details.
v1.0.11
- Improved clarity on user-facing device info: now only device name, state, and provider are displayed; internal infrastructure fields are hidden from users.
- Added privacy guidance: screenshots and UI data must only be shown to the user and never shared externally.
- Clarified error instructions and device reconnection steps to focus on user-relevant actions.
- Introduced a new section describing the two available control methods: direct Phone API and the autonomous Droidrun Agent, with advice on when to recommend each.
- Updated documentation links and organization for ease of reference; new explicit guidance on when to mention the Droidrun Agent.
- Removed mentions of non-supported tools and infrastructure details to simplify user experience and avoid confusion.
v1.0.10
## mobilerun 1.0.10
- Updated content in `subscription.md` (details on plans, pricing, credits, device types, or upgrade recommendations).
- No changes to the skill’s code or user-facing behaviors.
v1.0.9
- Updated and expanded the API reference details in api.md for more comprehensive usage guidance.
- Improved and clarified subscription and plan information in subscription.md, including details on credits, device types, and upgrade recommendations.
v1.0.8
mobilerun 1.0.8
- Updated setup instructions and troubleshooting guide in setup.md.
- No changes to code or functionality; documentation improvements only.
v1.0.7
- Added a "requires" metadata field specifying required environment variables and config keys (`MOBILERUN_API_KEY`, `skills.entries.mobilerun.apiKey`).
- No changes to user-facing functionality or documentation content.
v1.0.6
Updated description and cleaned up frontmatter
v1.0.5
Re-publish for security rescan
v1.0.4
Fix security scan: use openclaw config for API key persistence, remove shell profile writing
v1.0.3
Restructured from plugin to flat skill, added tags and category
v1.0.2
Add tags, category, and updated tagline
Metadata
Frequently Asked Questions
What is Mobilerun?
Load this skill whenever the user wants to control, automate, or interact with a phone or mobile device. This includes: tapping, swiping, typing, taking scre... It is an AI Agent Skill for Claude Code / OpenClaw, with 796 downloads so far.
How do I install Mobilerun?
Run "/install mobilerun" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Mobilerun free?
Yes, Mobilerun is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Mobilerun support?
Mobilerun is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Mobilerun?
It is built and maintained by johnmalek312 (@johnmalek312); the current version is v2.1.0.
More Skills