AgentDeck — AI skill for Claude Code
Mission control for AI coding agents on your Stream Deck Mini.
How to install AgentDeck
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open darkmatter2222/AgentDeck and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What AgentDeck does
Mission control for AI coding agents on your Stream Deck Mini. Monitor six sessions with live status indicators and one-touch window switching. Includes adapters for OpenCode, Claude Code, GitHub Copilot, Gemini CLI, and Cursor CLI.
Alternatives in AI
- Cc Mini — Ultra-light Harness scaffolding for AI agents, a mini version of claude code 825 ★
- Clother — Use multiple Claude Code–compatible LLM providers from one CLI, switching profiles instantly with simple cloth 430 ★
- Coffee CLI — A clean AI terminal workspace 180 ★
README
AgentDeck
**Six Stream Deck Mini keys for your coding agents: see activity, then press a key to focus the right window.**
AgentDeck connects OpenCode, Claude Code, GitHub Copilot CLI, Copilot in VS Code, Gemini CLI and Cursor CLI to one local controller. Each managed launch gets a stable key. A seventh launch waits for a vacancy; closing one does not shuffle the others. The controller uses direct USB HID, with no Elgato plugin or MCP server.

The preview shows every state together. In normal use, READY appears only when no agents are registered. Windows hardware and native harness acceptance gates are tracked in [the verification record](docs/TEST-RESULTS.md).
Demo video
[](https://www.youtube.com/watch?v=NTWLbLbJiO0)
[Watch on YouTube](https://www.youtube.com/watch?v=NTWLbLbJiO0). The original demo shows the OpenCode workflow; additional adapters have the coverage described below.

Choose your harness
| Harness | Integration / launch | Pending-input coverage |
|---|---|---|
| OpenCode | Global server plugin; opencode shim or oc |
Permission and structured-question IDs; SDK reconciliation when available |
| Claude Code | Project hooks; Launch-Claude.bat |
Identified AskUserQuestion calls; permission hooks show unknown instead of an invented count |
| GitHub Copilot CLI | Project hooks; Launch-Copilot.bat |
Activity only; approvals may still appear running |
| Copilot in VS Code | Project hooks; Launch-Copilot-VSCode.bat |
Activity only; isolated editor profile, preview integration |
| Gemini CLI | Project hooks; Launch-Gemini.bat |
Activity; recognized permission notifications show unknown |
Cursor CLI (agent) |
Project hooks; Launch-Cursor.bat |
Activity only; Cursor desktop integration is not included |
All five added hook adapters are imp
Related Skills
Herdeck
Control panel for AI coding agents running under herdr — on a hardware Stream Deck (Ulanzi D200 or Elgato), a
Claude Wall
Mission control for Claude Code — monitor, orchestrate, and interact with all your AI coding agents from one d
Comandos
Local-first terminal mission control for Claude Code, OpenAI Codex CLI, Gemini CLI and OpenCode. Run AI coding
AgentBar
Approve Claude Code permission prompts right from your menu bar — no terminal switching — plus live status for
Tmux Agent Indicator
Tmux plugin that gives visual feedback for AI agent states (running/needs-input/done). Supports Claude Code, C
Claude Code Status Line
A custom status line script for Claude Code that displays context window usage, git status, and model informat
Related Agents
Tp Migration Scout
PROACTIVELY use this BEFORE the user starts any migration — replacing an API, upgrading a framework version, r
Context Gathering
Use when creating a new task OR when starting/switching to a task that lacks a context manifest. ALWAYS provid
Unity Build Runner
Configures and triggers Unity builds via MCP. Handles platform switching, player settings, build profiles, Add