Skip to content

PointerSense v0.2.0

Choose a tag to compare

@Anuragh33 Anuragh33 released this 04 Jun 12:16
· 2 commits to main since this release

What's new in v0.2.0

A meaningful refresh — redesigned setup UI, two new AI providers, and detection bug fixes.

Setup window — full redesign

  • macOS-style sidebar layout: Overview / AI Providers / Detection / Output / Privacy
  • Indigo + lavender color palette matching the website
  • API Keys completely reworked — one clean row per provider with show/hide toggle, status badge, Save/Clear, and "Use this" to switch active provider
  • Segmented controls for detection mode, trigger mode, output mode, and answer length
  • iOS-style toggle switches for History and Launch-at-login
  • Cleaner Overview tab

New AI providers

  • Groq — default model openai/gpt-oss-20b (~1000 tokens/sec, the fastest cloud LLM available)
  • Command Code AI — OpenAI-compatible endpoint
  • Six total providers now: OpenAI, Anthropic, Gemini, Ollama Cloud, Groq, Command Code

Ollama Cloud model list refreshed

  • Added: Gemini 3 Flash (new default), Nemotron 3 Nano 4B, RNJ-1, Gemma 4 31B Cloud
  • Removed: nemotron-3-super and deepseek-v3.2 (these are reasoning models that caused 3-4s delays)
  • Fixed: qwen3.6 (didn't exist) → qwen3.5 (correct tag)

Detection bug fixes

  • Selected Text Only now works in Safari, Chrome, Notes, Electron apps. Previously it only read from the system-wide focused element which fails in webviews; now falls back to reading AXSelectedText from the element directly under the cursor.
  • Hold Option now feels instant — previously it still waited for the 700ms dwell delay, now it bypasses dwell entirely and uses a 600ms rate limit. Hold Option → fires immediately.

Requirements

  • macOS (Apple Silicon — aarch64)
  • Accessibility permission required on first launch
  • API key for at least one supported AI provider

Install

  1. Open PointerSense_0.2.0_aarch64.dmg and drag PointerSense to Applications.
  2. If macOS warns about an unidentified developer, right-click the app and choose Open.
  3. Grant Accessibility in System Settings → Privacy & Security → Accessibility.
  4. Open Setup from the menu bar, go to AI Providers, and paste your API key.