# omnimind-ai/OmniBot

Your on-phone / mobile AI Agent / Claw, capable of operating terminals and performing a wide range of tasks in the Android world || 你的手机 AI 代理，她可以操作终端，也可以完成 Android 世界的广泛任务

Repository: https://github.com/omnimind-ai/OmniBot
Canonical: https://ross.abutalabs.com/products/omnibot
Homepage: https://omnimind-ai.github.io/OmniBot-Docs/
Language: Dart
License: NOASSERTION
License Family: other
Topics: agent, ai, android, linux, skills, alpine, bot, gui, automation, on-device, claw, codex
Last push: 2026-08-25T04:10:28+00:00

## Health v2 (maintenance only)
Score: 82/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 99, longevity 12
- inputs: {"age_days": 168, "days_push": 8, "days_rel": 8, "gap_med": 1, "n_releases_24m": 58}
- flags: young, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1923, forks 135 (observed 2026-08-28T04:05:55.365334+00:00)

## What it is
OpenOmniBot (OmniBot) is an on-device AI agent app for Android, built with a native Kotlin host and Flutter UI, that runs a full understand-decide-execute-reflect loop. It combines chat, skills, an Alpine terminal workspace, MCP tool support, memory with embeddings, and Android system-level automation via accessibility and MediaProjection.

## Use cases
- run an AI agent directly on my Android phone
- automate tasks on Android with an AI assistant
- operate a terminal on my phone with AI
- use local LLMs on-device without cloud
- connect MCP tools to a mobile AI agent
- schedule alarms and calendar events via AI chat
- give an AI assistant file and browser access on mobile

## When to choose
- you want an open-source, self-contained AI agent running on your Android device
- you need system-level automation (accessibility, alarms, calendar, screen capture) driven by an LLM
- you want to mix local models (llama.cpp, MNN, QNN) with OpenAI/Anthropic-compatible remote models
- you want an extensible agent with skills, MCP servers, and a terminal workspace on mobile

## When to avoid
- you need an iOS or desktop agent - this is Android-only
- you want a simple chatbot without device automation or tool execution
- you require a permissively licensed library to embed in your own app (license is non-standard)
- you need a server-side or multi-user agent platform

## Facets
- artifact type: application
- maturity: active
- function: agent-framework, chatbot, mcp, llm-inference, workflow-automation, cli, gui
- domain: mobile-development, android-tools, large-language-models
- platform: cross-platform
- tags: on-device-ai, android-agent, flutter, accessibility-services, local-models, terminal-automation, mobile-assistant, ai-agents, automation, android, mobile

## Member repositories
- omnimind-ai/OmniBot (main) score 82

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:55.365334+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T03:09:38.842951+00:00, confidence not recorded.
  - readme: https://github.com/omnimind-ai/OmniBot (fetched 2026-08-28T04:05:55.365334+00:00, sha ff008bba0034)
  - homepage: https://omnimind-ai.github.io/OmniBot-Docs/ (fetched 2026-08-29T10:48:58.103209+00:00, sha e8af10e99021)
- Data as of 2026-08-30T08:39:29.467469+00:00.