AI Agent
Pushary
Pushary is a control panel for AI agents that lets you approve risky commands from your phone, set permission policies, and keep an audit trail.
Pushary
What is Pushary?
Pushary is a control panel for AI agents that allows you to manage permissions, approve commands from your phone, and maintain an audit trail across Claude Code, Codex, Cursor, and Hermes.
How to use Pushary?
- 1Run the one-command setup for your agent.
- 2Configure MCP server, permission hooks, event tracking, and API key.
- 3When your agent asks for a permission, your phone buzzes.
- 4Tap approve, deny, or choose an option from your lock screen.
- 5The agent continues without waiting.
Pushary Key Features
- One-command setup
- Auto-approve safe tools
- Escalate dangerous commands to your phone
- Per-tool timeout rules
- Live status and pending approvals
- Immutable audit trail
- Works with Claude Code, Codex, Cursor, and Hermes
Pushary Use Cases
- Run long refactors hands-free
- Manage multi-tool agent approvals remotely
- Keep audit trail of all decisions
Pushary Pricing & Free Credits
Pushary currently operates on a Free, Free Trial, Paid model.
Pushary Pros & Cons
Pros
- Hands-off operation while maintaining control
- Simple one-command setup
- Mobile notifications for approvals
- Compatible with popular AI coding agents
- Audit trail for all decisions
Cons
- Requires phone for approvals
- Limited to supported agents (Claude Code, Codex, Cursor, Hermes)
- Paid subscription after trial ends
What is Pushary best for?
- Developers using AI coding agents who want to run tasks hands-free without terminal babysitting