aaronnat23/disp8ch/code-dispatch
Claude Code and agent skill by aaronnat23/disp8ch
- community
Checking that the SKILL.md file is still in the repository.
Install
npx vskill@latest install aaronnat23/disp8ch/code-dispatchSKILL.md
Code Dispatch (ACP-style Coding Agent Delegation)
Delegate coding, file-system, and research tasks to a real CLI agent running on the user's machine. The orchestrator stays clean — no token overhead, no context window bloat. The specialist executes.
When to use
- Complex coding tasks that require multi-file edits, running tests, or reading the repo
- File-system operations beyond simple read/write (refactors, scaffolding, migrations)
- Research tasks that benefit from live CLI tools (git, grep, compilers, package managers)
- Any task where Claude Code's local tools + MCP servers give better results than inline agent calls
Rules
Agent ID: Always agent: "claude" (not "claude-code"). Use "gemini" or "codex" only when explicitly requested.
Mode selection by channel:
- Discord:
mode: "session"+thread: truefor multi-step work;mode: "run"for one-shot - Telegram / WhatsApp / Slack / Teams / BlueBubbles / WebChat: Always
mode: "run"— async thread binding is not available on these channels
Permission mode:
permission_mode: "approve-reads"(default): allows reads, prompts for writes/exec — safe for most taskspermission_mode: "approve-all": auto-approves all tool calls — only use for trusted, well-defined taskspermission_mode: "deny-all": read-only (Read, Glob, Grep, LS only) — use for audits or safe exploration
Timeout: Always set timeout_seconds (default 120, max 300). For complex refactors use 300.
Never nest: Do not call sessions_spawn from within a spawned session. The orchestrator dispatches; the specialist does the work.
Working directory: Always specify cwd pointing to the relevant project directory.
Example — Discord multi-step session
Example — one-shot on any channel
After dispatch
- Confirm task receipt to the user: "Dispatched to Claude Code — working on it."
- On
mode: "run": relay the result directly in your reply. - On
mode: "session"+ Discord: tell the user they can continue in the created thread. - On error: summarize the failure and suggest breaking the task into smaller steps.
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/code-dispatch/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/aaronnat23/disp8ch/code-dispatch)