ShizuStore

小万

omnimind-ai

0.7.0 · GitHub

Download APK
AI agents Android 10+ 2 weeks ago GPL-3.0
337 ShizuStore
18.4k GitHub
2.1k Stars
38 MB Size

More about this app

On-device AI agent with terminal, web browsing, device control, and system integration

English | ็ฎ€ไฝ“ไธญๆ–‡

Your On-Device AI Assistant

GitHub Repo stars GitHub Release
omnimind-ai%2FOmniBot | Trendshift

| Demo | Quick Start | Release | Issues |

OmniBot runs directly on your Android device and combines chat, agent tools, local workspaces, and system integrations in one app. ๐ŸŽ‰ Our iOS & macOS version: ViaVera

OmniBot is an on-device AI agent built with native Android Kotlin and Flutter. Instead of stopping at chat, it focuses on the full loop of understand -> decide -> execute -> reflect.

Core Capabilities

  • Extensible tool ecosystem: Skills, Alpine environment, browser access, MCP, and Android system-level tools.
  • System-level actions: Supports scheduled tasks, alarms, calendar creation/query/update, and audio playback control.
  • Personal context: Query authorized calendars, contacts, notifications, app usage, location, media, documents, SMS and call history without root, and manage calendar events and contacts.
  • Memory system: Short-term and long-term memory with embedding support.
  • Productivity tools: Read and write files, browse the workspace, use the browser, and access the terminal.

Example

Quick Start

Configure the app

Open the settings page from the left sidebar:

Configure AI capabilities Configure AI providers

Then open the scenario model settings:

Configure AI models

Note: Memory embedding requires an embedding model. For the best overall experience, the other scenarios should use multimodal or vision-capable models whenever possible.

Alpine environment

The app usually initializes the Alpine environment automatically on startup, and you can also manage that environment from the same settings area.

Native Gemini connection

Select Gemini in model provider settings, enter an API key, and keep the default base URL https://generativelanguage.googleapis.com/v1. Select a model for the built-in Agent. This connection uses the Interactions API for streamed replies, images, thinking levels, and local/MCP tools. External Agents require a Provider using their supported protocol.

Interactions use store=false. Original model steps and signatures are saved in local conversation history for subsequent turns and reopening a conversation. Older history without signatures is supplied as text. Context compaction preserves the active tool cycle. Google Search, Deep Research, and server-side background tasks are not enabled by this connection.

Use Cases

Personal context

Enable individual sources in Settings โ†’ App permissions โ†’ Personal context. AI queries shared device tools on demand; results may be sent to your configured model and are not automatically added to long-term memory. Calendar and contact read/write permissions are separate. Background location needs its own โ€œAllow all the timeโ€ grant. Approximate location and selected-only photo access are supported.

Notification access is separate from receiving Omnibot notifications. Observed notifications are cached locally for up to 7 days and 10,000 records, excluded from system backups, and can be cleared manually. Previously dismissed notifications and system-redacted content cannot be recovered. Revocation stops recording and querying without deleting conversations.

File queries return metadata first; requested contents are imported into the existing file inbox. Files and folders can be selected and access removed through the system-picker entry. SMS and call history are read-only and depend on installer/system eligibility; Omnibot does not request default SMS, phone or assistant roles. These features require neither root, Shizuku nor accessibility and do not access other apps' private databases.

Device MCP and the built-in Agent share tool definitions and execution. context_capabilities reports actual access; missing permissions do not open dialogs or terminate ordinary chat. Contact mutations use rawContactId and version from queries and affect only the selected account record. Calendar queries accept an offset; all-day dates use UTC and an exclusive end date.

Optional Root and device controls

Direct Root access is off by default. Enable Root access under Settings โ†’ App permissions โ†’ Advanced Access to request authorization from your device's Root manager. When available, direct Root takes priority over Shizuku; disabling it stops its commands and persistent shells. Shizuku remains independently available.

The agent can query and change supported system toggles, display and sound settings, manage app permissions, run diagnostics, and execute privileged shell commands. Mutations and shell commands require ACP client confirmation. State changes are checked against the device before being reported as successful; unsupported ROM interfaces return an explicit error. See Android privileged actions for the tool contract and compatibility details.

Skills

You can ask OmniBot to install a skill by simply sending it the repository link. Recommended collection: https://github.com/OpenMinis/MinisSkills

Enable or disable skills from the skill repository:

Skill store Skill example

Scheduled tasks

Scheduled task Timing

Scheduled tasks execute subagent flows. Alarms are reminder-only. A subagent can be assigned a complete task and behaves like a full agent.

Browser

Browser

Workspace

Workspace

Remote Codex bridge

To use Codex mode with Codex running on a PC or Mac, start codex-bridge on the computer where the Codex CLI is installed and logged in:

npx @thuocean/codex-bridge

Choose the LAN address and token mode in the terminal setup UI, then scan the printed QR code from OpenOmniBot's Codex settings. For advanced options and troubleshooting, see the codex-bridge README.

Development Guide

Requirements

  • Flutter SDK 3.47.2+
  • JDK 17+
  • Node.js 20.19+ or 22.12+ and pnpm 10.28.0 (for WebUI development)

Get the code

git clone https://github.com/omnimind-ai/OpenOmniBot.git
cd OpenOmniBot

cd ui
flutter pub get

If Flutter reports Could not read script '.../ui/.android/include_flutter.groovy', run:

flutter clean
flutter pub get

Develop the WebUI locally

The WebUI in webchat/ is a standalone React + TypeScript + Vite project. During local development, Vite serves the frontend with hot reload and proxies /webchat/api requests to the Android app's local service.

  1. Install and start OpenOmniBot on an Android device. Keep the computer and device on the same trusted LAN.
  2. In the app, open Settings > Local Service, enable the service, and copy its address and Token. The default port is 8899, but always use the address shown by the app.
  3. Start the WebUI development server from the repository root, replacing the sample address with the Android local-service address (do not append /webchat):
cd webchat
pnpm install --frozen-lockfile

VITE_WEBCHAT_PROXY_TARGET=http://192.168.1.20:8899 pnpm dev

On PowerShell, set the proxy target first:

$env:VITE_WEBCHAT_PROXY_TARGET = "http://192.168.1.20:8899"
pnpm dev

Open the URL printed by Vite (normally http://localhost:5173) and enter the Token copied from the app. Use pnpm dev for end-to-end API/SSE testing; the proxy keeps session cookies, realtime events, workspace access, and browser mirroring on the same local origin.

Before submitting WebUI changes, run:

cd webchat
pnpm run typecheck
pnpm run build

The production files are generated in webchat/dist/. Both dist/ and node_modules/ are local outputs and must not be committed.

Android builds handle the WebUI automatically: Gradle runs the locked pnpm install, executes the Vite production build, clears stale WebChat assets, and copies only dist/ into the APK. To verify this step without building the full app, run:

./gradlew :app:syncWebChatBundle -Ptarget=lib/main_standard.dart

Flutter Web is not part of this workflow.

Build and install

Android builds use OMNIBOT_UPDATE_WORKER_URL for the website backend's update checks, anonymous statistics, cloud policy, and official VLM proxy configuration. It defaults to https://omnibot.omnimind.com.cn; upstream proxy credentials stay on the backend. See the integration notes. Debug APK builds use the OpenAI-compatible LLM API configured by LLMTHU_API_BASE, LLMTHU_API_KEY, and LLMTHU_MODEL for normal LLM requests, context compaction, and scene.vlm.operation.primary. For local acceptance testing, Release builds can explicitly enable -POOB_BUNDLE_LLMTHU_PROVIDER=1 to bundle the same LLMTHU_* configuration for use out of the box. Regular Release and CI builds do not bundle this configuration by default.

cd ..

./gradlew :app:installDevelopStandardDebug -Ptarget=lib/main_standard.dart

For acceptance testing on a physical device, use the repository's LLMTHU installation script. It enables the LLMTHU provider by default and reads the API key from LLMTHU_API_KEY in the current shell. If the key is missing, the script fails immediately instead of producing an unusable test APK. Regular Release and CI builds still do not automatically bundle the key.

export LLMTHU_API_KEY='your-api-key'
./scripts/install_release_llmthu_device.sh

Architecture Overview

OpenOmniBot/
โ”œโ”€โ”€ app/                        # Android host app: entry point, agent orchestration, system abilities, MCP, services
โ”œโ”€โ”€ ui/                         # Flutter Android UI: chat, settings, tasks, and memory
โ”œโ”€โ”€ webchat/                    # React + TypeScript WebUI; Vite builds the static bundle packaged by Android
โ”œโ”€โ”€ baselib/                    # Shared core libraries: database, storage, networking, model config, permissions
โ”œโ”€โ”€ assists/                    # Shared task lifecycle and chat/model coordination
โ”œโ”€โ”€ uikit/                      # Native overlay UI: floating ball, overlay panels, half-screen surfaces
โ””โ”€โ”€ ReTerminal/core/            # Embedded terminal experience modules

More

Thanks to developers from LINUX.DO and other communities for supporting OpenOmniBot.

Special thanks to these open-source projects:

WeChat Group
WeChat Group
Join the Discord community
Close

How Shizuku is used

Can run AI-directed privileged shell, app, settings and device actions via `Shizuku` user service shell commands.

This is an AI-assisted analysis of Shizuku-related usages in the app's public source code. It is best effort, so it may not catch every single usage.

How this app uses Shizuku

The on-device AI agent uses Shizuku to run privileged Android operations that normal apps cannot perform.

  • Run privileged shell commands: the agent can run a one-shot shell command with a chosen working directory and environment through the privileged process, with dangerous commands blocked by policy.
  • Use persistent shell sessions: the agent can start a reusable privileged shell session, run commands in it, read back output transcripts and stop the session when done.
  • Launch and stop apps: the agent can launch an app or a specific app screen with am start or the launcher trigger, and force stop other apps with am force-stop.
  • Manage app permissions: the agent can grant or revoke an app permission with pm grant and pm revoke, and set an app operation mode with appops set.
  • Read and change settings: the agent can read system properties and settings values, and write system, secure or global settings values.
  • Control device features: the agent can send key events, expand notifications or quick settings, and turn WiFi or mobile data on or off.
  • Inspect system diagnostics: the agent can list installed packages, read selected system service status and read recent logs.

Android APIs or commands used

  • ProcessBuilder
  • /system/bin/sh -lc
  • am start
  • am force-stop
  • pm grant
  • pm revoke
  • pm list packages
  • appops set
  • settings get
  • settings put
  • input keyevent
  • cmd statusbar expand-notifications
  • cmd statusbar expand-settings
  • svc wifi enable
  • svc wifi disable
  • svc data enable
  • svc data disable
  • getprop
  • dumpsys
  • logcat
  • monkey -p

Notable details

Privileged agent actions ask for explicit user confirmation before running and show the pending command first; unconfirmed or denied actions do not run. A built-in safety policy blocks reboot, power off, recovery wipe, repartition, fastboot, block device and protected system partition writes. Mobile data control and kernel log access are only available in some Shizuku configurations.

Close

Changelog

What's new for version 0.7.0

Close

Permissions

47 permissions requested

  • android.permission.CAMERA
  • android.permission.BLUETOOTH
  • android.permission.BLUETOOTH_ADMIN
  • android.permission.BLUETOOTH_CONNECT
  • android.permission.BLUETOOTH_SCAN
  • android.permission.CHANGE_WIFI_STATE
  • android.permission.MODIFY_AUDIO_SETTINGS
  • android.permission.WRITE_SETTINGS
  • android.permission.INTERNET
  • android.permission.ACCESS_WIFI_STATE
  • android.permission.ACCESS_NETWORK_STATE
  • android.permission.CHANGE_NETWORK_STATE
  • android.permission.POST_NOTIFICATIONS
  • android.permission.REQUEST_IGNORE_BATTERY_OPTIMIZATIONS
  • android.permission.WAKE_LOCK
  • android.permission.QUERY_ALL_PACKAGES
  • android.permission.FOREGROUND_SERVICE
  • android.permission.FOREGROUND_SERVICE_DATA_SYNC
  • android.permission.SYSTEM_ALERT_WINDOW
  • android.permission.BACKGROUND_ACTIVITY_STARTS
  • android.permission.READ_MEDIA_IMAGES
  • android.permission.READ_MEDIA_AUDIO
  • android.permission.VIBRATE
  • android.permission.VIBRATE_ALWAYS
  • android.permission.WRITE_EXTERNAL_STORAGE
  • android.permission.READ_EXTERNAL_STORAGE
  • android.permission.MANAGE_EXTERNAL_STORAGE
  • android.permission.REQUEST_INSTALL_PACKAGES
  • android.permission.SCHEDULE_EXACT_ALARM
  • com.android.alarm.permission.SET_ALARM
  • android.permission.RECEIVE_BOOT_COMPLETED
  • android.permission.READ_CALENDAR
  • android.permission.WRITE_CALENDAR
  • android.permission.READ_CONTACTS
  • android.permission.WRITE_CONTACTS
  • android.permission.READ_SMS
  • android.permission.READ_CALL_LOG
  • android.permission.PACKAGE_USAGE_STATS
  • android.permission.ACCESS_COARSE_LOCATION
  • android.permission.ACCESS_FINE_LOCATION
  • android.permission.ACCESS_BACKGROUND_LOCATION
  • android.permission.READ_MEDIA_VIDEO
  • android.permission.READ_MEDIA_VISUAL_USER_SELECTED
  • android.permission.FOREGROUND_SERVICE_MEDIA_PLAYBACK
  • android.permission.FOREGROUND_SERVICE_SPECIAL_USE
  • cn.com.omnimind.bot.DYNAMIC_RECEIVER_NOT_EXPORTED_PERMISSION
  • moe.shizuku.manager.permission.API_V23
Close

Sources

2 sources

0.7.0 arm64-v8a GitHub ยท cn.com.omnimind.bot
Selected Download
0.6.3 Universal GitHub ยท cn.com.omnimind.bot
Download