avgvstvs96/dotfiles/tmux
Claude Code and agent skill by avgvstvs96/dotfiles
- community
Checking that the SKILL.md file is still in the repository.
Control tmux from the CLI for session/window/pane inspection, command execution, capture, and troubleshooting. Use when the user asks to work in a specific tmux target (pane id like %232, window name/index, or session name), run commands in existing panes, validate pane output, or automate interactive terminal workflows.
Install
npx vskill@latest install avgvstvs96/dotfiles/tmuxSKILL.md
tmux CLI Control
Use tmux as a remote terminal control plane: inspect targets, send input, verify output, and keep work scoped to the pane/session the user names.
Core Workflow
- Resolve the target the user provided.
- Inspect the target state before sending input.
- Send input with explicit Enter handling.
- Capture pane output and verify the action happened.
- Report what was run and what changed.
Resolve Targets
Accept user target hints in these forms:
- Pane id:
%232 - Session name:
workspace - Window selector:
workspace:6orworkspace:6.2 - Window name/title hints (resolve via listing)
Use:
If the user gave a session/window name instead of a pane id, resolve it to a concrete pane id first.
Inspect and Capture
Before sending commands, quickly inspect context:
Use capture-pane again after actions to verify outcomes.
Send Input Reliably
For normal shell commands:
For arbitrary text (safer quoting):
For multiline payloads, prefer tmux buffers:
Critical Gotcha: Prompt Not Actually Sent
Interactive agents often leave prompt text in the input line when only text is injected and Enter is not sent correctly. Never assume the prompt executed just because text appears in the pane.
Always do all three steps:
- Insert prompt text (
send-keys -lorpaste-buffer). - Send Enter separately with
tmux send-keys -t <target> C-m. - Verify by capturing pane output; if prompt is still sitting in input, send
C-magain and re-check.
Safety and Accuracy
- Prefer targeting a specific pane id once resolved.
- Echo or capture output after each important action.
- Keep commands idempotent when possible during testing.
- If multiple panes are plausible, list options and pick the one matching user intent.
Interactive Coding Agents
For agent-specific launch commands, prompt-delivery patterns, and verification tactics, read:
references/coding-agents.md
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingclaude/.claude/skills/tmux/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/avgvstvs96/dotfiles/tmux)