Unclaimed project

Are you a maintainer of droidclaw? Claim this project to take control of your public changelog and roadmap.

Claim this project

Changelog

droidclaw

turn old phones into ai agents - give it a goal in plain english. it reads the screen, thinks about what to do, taps and types via adb, and repeats until the job is done.

unitedbyai/droidclaw

Back to changelog

NewFebruary 23, 2026

v0.4.0

Fix: correct dashboard URL to app.droidclaw.ai and make it clickable
add screenshot action using GLOBAL_ACTION_TAKE_SCREENSHOT
show LLM setup banner on devices page when unconfigured
Fix: use per-route auth middleware on pairing routes
replace API key onboarding with OTP pairing + self-host toggle
replace free-text model input with provider-specific dropdown
add Pair Device button and OTP modal to devices page
add pairing API client for OTP flow
add OTP pairing create/claim/status endpoints
add pairing_code table for OTP device pairing
Refactor: separate voice transcription key from user LLM config
add setup checklist and example goals for onboarding
Redesign: Google Account-style card layout with Solar icons
Refactor: use SvelteKit enhance() for create and delete forms
Refactor: use SvelteKit enhance() for optimistic config update
Fix: duplicate devices, content width, and settings refresh
Fix: reset goal status to idle after completed/failed
Fix: move tap handling from Compose to touch listener
Fix: restore drag by consuming ACTION_DOWN in touch listener
Refactor: route pill taps through Compose callbacks instead of touch listener
Fix: address code review feedback on pill widget
Refactor: rewrite pill widget from circle to text-based pill
auto-connect, assistant invocation, suggestion cards, and onboarding assistant step
add overlay command panel, dismiss target, vignette, voice integration, and theme updates
Fix: address critical review issues in voice overlay
add RECORD_AUDIO runtime permission handling
wire voice recording and transcript into ConnectionService
expand AgentOverlay with voice mode state machine
add voice overlay UI with transcript and action buttons
add animated gradient border composable
add VoiceRecorder with AudioRecord PCM streaming
add voice protocol models and overlay mode enum
wire voice messages into device handler
add voice session handler with Groq Whisper STT
add voice overlay protocol types
make battery optimization permission optional
replace get started button with open dashboard link
add APK download links and launch banner across site, web, and README
Fix: update email sender name and switch to production Polar checkout link
Fix: handle Polar activation limit gracefully + switch checkout to command pattern