Index / DroidClaw · updated Oct 2, 2026
DroidClaw
unitedbyai/droidclaw · healthy · rank 23 of 43 by stars
DroidClaw turns old Android phones into autonomous AI agents that read the screen via accessibility trees and tap, type, and swipe through ADB until a plain-English goal is done. Its killer trick is delegating tasks to on-device apps like ChatGPT or Gemini — no API keys needed, it just uses apps like a human would.
Facts
Repository
Runtime
Posture
Community
Security breakdown
Composite 45 / 100 · how these are scored
higher is safer
higher is safer
higher is safer
higher is safer
higher is riskier
Model access
Read from the repository, not written by a model · 29 files examined
3 providers · runs locally · gateway support
Pinned models
llama-3.3-70b-versatile released 2024-12-06 — 22 months old (from the public model catalogue)
Pin last edited 8 months ago
Several defaults are pinned; the project picks one per provider rather than shipping a single default.
Evidence
Decision
Why choose DroidClaw over OpenClaw?
Why choose this
- Controls real Android UIs via accessibility trees and ADB
- Uses installed apps like ChatGPT/Gemini without API keys
- Gives new life to spare/old Android hardware
Tradeoffs
- Narrower scope — Android UI automation only
- Early-stage (v0.5.x) with active bug fixing
- No documented sandboxing or permission model
Best fit
- Repurposing old Android phones as AI agents
- Automating mobile apps without APIs or integrations
- Delegating tasks to on-device ChatGPT/Gemini apps
Avoid if
- You need desktop or server-side automation
- You require strong sandboxing around agent actions
- You want a mature, stable production platform
Evidence is solid on architecture and activity from the README and recent commits, but there is zero Reddit discussion and web sentiment is mostly about the broader Claw ecosystem rather than DroidClaw itself. Security posture and license are uncertain due to missing documentation.
AI layer reviewed Sep 21, 2026 · how this is written
Star activity
1,566 stars today
Overview
DroidClaw is an open-source AI agent that turns Android phones — including old, spare devices — into autonomous operators. You give it a goal in plain English, and it runs a perception → reasoning → action loop: it dumps the accessibility tree via ADB, parses the XML into interactive UI elements, diffs against the previous screen to detect changes, optionally captures a screenshot, then decides whether to tap, type, swipe, or launch an app. It repeats this cycle (up to a step budget) until the goal is complete.
The standout capability is API-free delegation: DroidClaw can open the ChatGPT, Gemini, or Google Search apps on the device itself, submit a query, and bring the result back — using the apps exactly as a human would, with no API keys required. The project ships an Android APK (currently v0.5.3), a web dashboard, scheduled goals via QStash, and a documented set of 26 actions with failure recovery and a pairing flow.
Compared to OpenClaw, DroidClaw is a specialist rather than a generalist: it focuses entirely on mobile UI automation rather than being a broad personal assistant platform. Recent commits show active development around session management, scheduled goals, and dashboard reliability, though the project is still pre-1.0 with visible bug-fixing churn. Security hardening (sandboxing, permission scoping) is not prominently documented, so it carries meaningful operational risk given it has full control of a device.