droidrun-docs
This skill provides documentation on the Mobilerun/DroidRun framework regarding agent orchestration, CLI commands, and mobile automation.
Install
mkdir -p .claude/skills/droidrun-docs && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/7577" && unzip -o skill.zip -d .claude/skills/droidrun-docs && rm skill.zipInstalls to .claude/skills/droidrun-docs
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Mobilerun documentation reference. Use when users ask about Mobilerun setup, configuration, SDK usage, CLI commands, device setup, agents, architecture, app cards, credentials, tracing, Docker, migration, structured output, or any Mobilerun "how do I..." questions.Key capabilities
- →Orchestrate mobile agents for automation
- →Execute atomic actions like click and swipe
- →Parse structured output from mobile devices
- →Manage credentials via YAML
- →Trace automation flows with Phoenix
How it works
It uses a multi-agent architecture to plan and execute mobile device actions based on natural language input.
Inputs & outputs
When to use droidrun-docs
- →Setting up mobile automation agents
- →Configuring DroidRun SDK
- →Debugging mobile automation flows
- →Parsing structured output from devices
About this skill
Mobilerun
Mobilerun is an open-source (MIT) framework for controlling Android and iOS devices through LLM agents. It enables mobile automation using natural language commands.
- GitHub: https://github.com/droidrun/mobilerun
- Docs site: https://docs.mobilerun.ai
- License: MIT
- Install:
uv tool install mobilerun(Google Gemini, OpenAI, DeepSeek, Ollama, OpenRouter included by default) - Optional extras:
anthropic,langfuse - Requires: Python 3.11+, ADB, Portal APK on device
Architecture
Mobilerun uses a multi-agent architecture coordinated by MobileAgent:
- Reasoning mode (
reasoning=True): Manager plans → Executor acts → loop until done - Direct mode (
reasoning=False): CodeActAgent generates and executes Python code directly
Key agents: ManagerAgent (planning), ExecutorAgent (actions), CodeActAgent (direct execution), ScripterAgent (off-device computation), StructuredOutputAgent (typed data extraction).
Atomic actions available to agents: click, long_press, type, system_button, swipe,
open_app, get_state, take_screenshot, complete.
Repository Structure
Source code is at mobilerun/ (Python package). Key locations:
| Path | Description |
|---|---|
mobilerun/agent/droid/ | MobileAgent orchestrator |
mobilerun/agent/codeact/ | CodeActAgent (direct mode) |
mobilerun/agent/scripter/ | ScripterAgent (off-device scripts) |
mobilerun/agent/oneflows/ | StructuredOutputAgent |
mobilerun/agent/utils/ | Tools, executor, tracing, async utils |
mobilerun/tools/ | Device tools (ADB, iOS, portal client) |
mobilerun/cli/ | CLI entry point (click-based) |
mobilerun/config/prompts/ | Jinja2 prompt templates per agent |
mobilerun/config/app_cards/ | App-specific instruction cards |
mobilerun/credential_manager/ | YAML-based credential storage |
mobilerun/telemetry/ | Phoenix tracing integration |
Documentation
Read the relevant file(s) from docs/v5/ based on the user's question. Do not guess — always
read the doc before answering.
| Topic | File |
|---|---|
| Overview | docs/v5/overview.mdx |
| Quickstart | docs/v5/quickstart.mdx |
| Concepts | |
| Architecture & agents | docs/v5/concepts/architecture.mdx |
| Events & workflows | docs/v5/concepts/events-and-workflows.mdx |
| Prompts | docs/v5/concepts/prompts.mdx |
| Scripter agent | docs/v5/concepts/scripter-agent.mdx |
| Shared state | docs/v5/concepts/shared-state.mdx |
| Features | |
| App cards | docs/v5/features/app-cards.mdx |
| Credentials | docs/v5/features/credentials.mdx |
| Custom tools | docs/v5/features/custom-tools.mdx |
| Custom variables | docs/v5/features/custom-variables.mdx |
| Structured output | docs/v5/features/structured-output.mdx |
| Telemetry | docs/v5/features/telemetry.mdx |
| Tracing | docs/v5/features/tracing.mdx |
| Guides | |
| CLI usage | docs/v5/guides/cli.mdx |
| Device setup | docs/v5/guides/device-setup.mdx |
| Docker | docs/v5/guides/docker.mdx |
| Migration v3→v4 | docs/v5/guides/migration-v3-to-v4.mdx |
| SDK | |
| MobileAgent | docs/v5/sdk/droid-agent.mdx |
| ADB tools | docs/v5/sdk/adb-tools.mdx |
| iOS tools | docs/v5/sdk/ios-tools.mdx |
| Base tools | docs/v5/sdk/base-tools.mdx |
| Configuration | docs/v5/sdk/configuration.mdx |
| API reference | docs/v5/sdk/reference.mdx |
For deeper implementation details, check the source code directly in mobilerun/.
When not to use it
- →When Python 3.11+ is not available
- →When ADB or Portal APK is missing
Prerequisites
Limitations
- →Requires specific device setup
- →Limited to Android and iOS
How it compares
It enables natural language control of mobile devices using LLM agents, unlike traditional script-based automation.
Compared to similar skills
droidrun-docs side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| droidrun-docs (this skill) | 1 | 2mo | Review | Intermediate |
| codex-cli-bridge | 9 | 9mo | Review | Intermediate |
| mobile-use-setup | 1 | 6mo | Caution | Beginner |
| mflux-release | 1 | 5mo | No flags | Beginner |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
codex-cli-bridge
alirezarezvani
Bridge between Claude Code and OpenAI Codex CLI - generates AGENTS.md from CLAUDE.md, provides Codex CLI execution helpers, and enables seamless interoperability between both tools
mobile-use-setup
minitap-ai
Interactive setup wizard for Minitap mobile-use SDK. USE WHEN user wants to set up mobile automation, configure mobile-use SDK, connect iOS or Android devices, or create a new mobile testing project.
mflux-release
filipstrand
Prepare a release in mflux (version bump, changelog, uv lock) without tagging/publishing.
code-changelog
bear2u
AI가 만든 모든 코드 변경사항을 reviews 폴더에 기록하고 간단한 HTML 뷰어로 웹 브라우저에서 실시간 확인할 수 있습니다. 매 수정마다 문서가 생성되고 Python 서버로 즉시 확인 가능합니다.
skills
emergencescience
Instructions for AI coding agents (Claude Code, Hermes, Cursor, etc.) contributing to this project.
project-maintenance
Dilaansg
Use when changing the Farmadil System project, especially when adding new libraries, updating requirements.txt, or making major changes that require updating DESCRIPCION_SISTEMA.md.