Pushary

Remote agent supervision

Approve or deny AI agent requests from your phone while away from your desk.

Long-running tasks

Hand off a multi-step task and get notified only when a decision is needed, instead of babysitting the agent.

Deployment approvals

Tap approve or deny on production deployment requests sent to your phone or Slack.

Multi-agent management

Control requests from Claude Code, Codex, Cursor, Hermes, Windsurf, and any MCP agent from one place.

Workflow automation with human checkpoints

Pause an n8n workflow for human approval before continuing.

Slack-based approvals

Answer agent questions directly in Slack with Approve and Deny buttons, with results updating in the message.

One-tap approval

Approve or deny agent requests from your phone's lock screen with a single tap.

Multi-channel delivery

Permission requests reach you on your phone, Slack, the web app, and an audit trail simultaneously.

Cross-agent support

Works with Claude Code, Codex, Cursor, Hermes, Windsurf, and any MCP agent.

YOLO mode

Automatically approves all steps (including risky ones like rm and deploy) without pinging you.

Safe auto-run mode

Pushary automatically runs only safe steps and only pings you for decisions that need your call.

Slack integration

Agent questions appear in Slack with Approve and Deny buttons; one click answers and updates the message.

n8n integration

Add a human checkpoint node to any n8n workflow—the flow pauses, asks you, and continues with your answer.

Audit trail

Every permission request is logged in a full audit trail across all channels.

Three question types

Supports different kinds of agent requests (run a command, change a file, or pick between options).

One-command setup

Get Pushary running in under two minutes with a single command.