zellij-agent-control

Discover terminal panes, read output, and send keystrokes in zellij sessions.

Updated Apr 13, 2026
One-click install
npx skills add https://github.com/bermudi/skills --skill zellij-agent-control
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: zellij-agent-control
Source: https://github.com/bermudi/skills/tree/main/zellij-agent-control
Command: npx skills add https://github.com/bermudi/skills --skill zellij-agent-control

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automate the observation and control of terminal panes inside a zellij session to drive coding agents, monitor progress, and unblock workflows.

Core Features & Use Cases

  • Pane discovery and targeting: locate the agent pane within a session and direct commands to it.
  • Screen reading and decision making: dump pane output to determine next actions based on prompts and status.
  • Input automation: send keystrokes or text to the pane to control the agent, support wait loops, and coordinate multiple agents.
  • Use cases: automate repetitive agent interactions, unblock stalled agents, script cross-agent workflows, and respond to agent prompts through zellij.

Quick Start

Identify the agent pane and start driving it with scripted keystrokes.

Frequently Asked Questions about zellij-agent-control

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate terminal agent workflows in zellij?

Automating terminal agent workflows in zellij involves observing and controlling terminal panes to drive coding agents, monitor progress, and unblock stalled tasks using pane discovery, screen reading, and scripted keystrokes.

Can I send keystrokes to a specific zellij pane to control a coding agent?

Yes, you can send keystrokes or text to a specific zellij pane by discovering the target agent pane within the session and directing scripted input automation commands to control it.

What is the best way to unblock stalled coding agents in a terminal multiplexer?

The best way to unblock stalled coding agents in zellij is to dump pane output to read the screen, determine next actions based on prompts, and send necessary keystrokes to respond and resume the workflow.

How do I read terminal screen output to decide actions for an automated agent?

You can read terminal screen output to decide actions for an automated agent by dumping the pane output within a zellij session to monitor progress, determine next actions based on prompts, and coordinate responses.

Do I need any external dependencies to script cross-agent workflows with zellij?

No external dependencies are required to script cross-agent workflows with zellij. The automation relies entirely on built-in zellij action commands to discover panes, read output, decide actions, and send keystrokes.