Pushary - Control panel for AI agents

Hey PH! Aadil here, solo founder building Pushary. Here's the thing that bugs me: I run Claude Code on a big refactor, go make coffee, come back 40 minutes later — and the agent finished 35 minutes ago. Or worse, it's been waiting for my permission to run a command the entire time. So I built Pushary, push notifications for AI coding agents. How it works: 1. Run `npx @pushary/agent-hooks setup` 2. Pick your agent (Claude Code, Cursor, Codex, Windsurf, Hermes, Lovable) 3. Done. Your phone now buzzes when your agent finishes, fails, or needs approval. The permission hooks are the killer feature. Your agent wants to run `rm -rf ./src`? You get a push notification with Allow/Deny buttons. Tap from your lock screen. Agent continues. You never leave the couch. What's different from launch #2: - Permission hooks (approve/deny from your phone) - 3 question types (yes/no, multiple choice, free text) - Works with 6+ agents now (was just Cursor + Claude) - One unified CLI setup (was manual JSON config) Free to start. Would love your feedback, especially if you're running multiple agents and losing time to tab-checking. 🔗 pushary.com/ai-coding