deadeye636/switchboard
Desktop app to browse, search, launch and monitor your coding-CLI sessions (Claude Code, Codex, Hermes, Pi, Antigravity) across every project.
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code
$ npx degit deadeye636/switchboard/.claude ./rig-switchboard # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit deadeye636/switchboard/.agents/skills/release-notes .claude/skills/release-notes $ npx degit deadeye636/switchboard/.claude/skills/code-discipline .claude/skills/code-discipline $ npx degit deadeye636/switchboard/.claude/skills/output-contract .claude/skills/output-contract $ npx degit deadeye636/switchboard/.claude/skills/severity-gates .claude/skills/severity-gates
$ curl -fsSL --create-dirs -o .claude/agents/backend-format-expert.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/backend-format-expert.md $ curl -fsSL --create-dirs -o .claude/agents/electron-desktop-expert.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/electron-desktop-expert.md $ curl -fsSL --create-dirs -o .claude/agents/implementer.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/implementer.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/orchestrator.md $ curl -fsSL --create-dirs -o .claude/agents/performance-memory-expert.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/performance-memory-expert.md $ curl -fsSL --create-dirs -o .claude/agents/planner.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/planner.md $ curl -fsSL --create-dirs -o .claude/agents/terminal-pty-expert.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/terminal-pty-expert.md $ curl -fsSL --create-dirs -o .claude/agents/test-runner.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/test-runner.md $ curl -fsSL --create-dirs -o .claude/agents/ui-codemirror-expert.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/ui-codemirror-expert.md $ curl -fsSL --create-dirs -o .claude/agents/verifier.md https://raw.githubusercontent.com/deadeye636/switchboard/main/.claude/agents/verifier.md
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"hooks": {
"PreToolUse": [
{
"matcher": "Bash|PowerShell",
"hooks": [
{
"type": "command",
"command": "node"
}
]
}
]
}
} Skills (4)
Subagents (10)
| backend-format-expert model: sonnet | Domain advisor for AI CLI transcript formats (Claude, Codex, Pi, Hermes, agy), SQLite stores, Protobuf wire formats, and session lifecycle state machines. Read-only, with web research. |
| electron-desktop-expert model: sonnet | Domain advisor for Electron runtime, IPC bridge security, native module compilation (better-sqlite3, node-pty), multi-OS window lifecycle, and packaging. Read-only, with web research. |
| implementer model: sonnet | Surgical code implementer bound to exactly one plan step (T<n>) at a time. Makes the minimal change, runs the targeted tests, and stops after two failed fix attempts. Spawn only on the user's say-so; |
| orchestrator model: opus | Main-session agent only — start it with `claude --agent orchestrator`. Technical lead and dispatcher for a multi-step plan. Consults domain advisors, has the planner write plans and specs, and hands s |
| performance-memory-expert model: sonnet | Domain advisor for event-loop latency, worker thread delegation, SQLite query and index optimization, FTS5 full-text search, and memory leak prevention. Read-only, with web research. |
| planner model: sonnet | Solution architect. Turns an issue and the decisions around it into a spec (docs/specs/) and a phased plan bundle (docs/plans/<number>-<slug>/PLAN.md) with atomic step IDs. Writes only to docs/specs/, |
| terminal-pty-expert model: sonnet | Domain advisor for pseudoterminal (PTY) emulation, @xterm/xterm rendering, ConPTY integration, ANSI/VT escape sequence parsing, and terminal buffer management. Read-only, with web research. |
| test-runner model: haiku | Isolated test execution and log filtering agent. Use whenever running long, noisy test suites (e.g. `npm test`, targeted `node --test` runs, or `npm run backends:*` audits) to avoid polluting the main |
| ui-codemirror-expert model: sonnet | Domain advisor for Vanilla JS UI rendering (morphdom), CodeMirror 6 extensions and diff views, responsive pane layouts, keyboard navigation, and accessibility. Read-only, with web research. |
| verifier model: opus | Spec/acceptance conformance verifier. Use BEFORE closing a substantial issue — multi-file, security-sensitive, or one with explicit acceptance criteria — to independently check that the implementation |
Hooks (1)
| event | matcher | runs |
|---|---|---|
| PreToolUse | Bash|PowerShell | node |
Slash commands (1)
/release
Similar rigs
MValles02/claude-config
Personal Claude Code configuration (agents, hooks, settings)
Fort Knox 1.4k tok · · Q29
joshuakatt/Bob-The-Builder
Autonomous task orchestrator for Kiro + Claude specs.
YOLO Cowboy 1.4k tok · · Q0
kr4ckhe4d/local-llm-benchmarks
Measured llama.cpp benchmarks on AMD Radeon RDNA4 with ROCm: RX 9070 XT + Radeon AI PRO R9700 (48 GB). Qwen3.8, Gemma 4, GLM, gpt-oss; GGUF quant quality (KLD), MTP speculative decoding, 128K context, Claude Code with local models.
Minimalist 85 tok · · Q0
Yassinekrn/claude-setup
Personal Claude Code setup — orchestrator/subagent delegation policy, custom agents, and a colored PowerShell statusline (git branch, model+effort, tokens, cost, rate limits).
Orchestrator 1.2k tok · · Q15