from my-ai-tools75
Programmatically control tmux sessions to run interactive CLIs (REPLs, debuggers, databases), capture pane output, and automate terminal workflows safely.
This skill enables an agent to create, control, and capture output from tmux sessions to run interactive terminal applications without blocking the host terminal. It documents practical patterns: isolated sockets to avoid user conflicts, send-keys sequences, capture-pane usage, and session lifecycle management. It includes recipes for Python REPLs, GDB debugging, Node/psql sessions, and recommends LogPilot for automated output capture and AI analysis.
Use this skill when an agent must run interactive CLIs, test or debug programs, or orchestrate ephemeral terminal sessions as part of automation. It is especially useful when you need reproducible interaction scripts (send-keys → capture → parse) or when integrating terminal workflows into MCP servers and agent runtimes.
Works well with agents that can run shell commands and manage processes (Claude Code, Codex/Copilot-style editors, Cursor).
This skill has not been reviewed by our automated audit pipeline yet.
