delegating-to-agents
Delegate bounded work to other AI agents while preserving context, ownership, and progress checks.
- License 许可
- MIT
- 最近更新
- 2026-10-25
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- delegating-to-agents
- risk
- critical
- tags
- [ "agents", "delegation", "orchestration" ]
- tools
- [ "claude", "codex" ]
- author
- davidondrej
- source
- community
- license
- MIT
- category
- agent-orchestration
- date_added
- 2026-07-07
- description
- Delegate bounded work to other AI agents while preserving context, ownership, and progress checks.
- source_repo
- davidondrej/skills
- source_type
- community
- license_source
- https://github.com/davidondrej/skills/blob/main/LICENSE
03 SKILL.md 正文
Delegating to Agents
When to Use
- Use when work should be handed to another AI agent with a complete prompt and progress checks.
- Use when you need to relay instructions to terminal or TUI agents without losing context.
Which agent to pick
- Coding (default) → Codex CLI. Strongest coding agent, especially for complex, long-running SWE tasks. It's on an unlimited-usage plan — effectively unlimited, don't ration it.
- Most other tasks → Pi Agent (
piin a cmux terminal). All Pi agents run opus-4.8-fast via OpenRouter at xhigh reasoning effort. - Frontend / design → Pi. Opus 4.8 Fast beats Codex on UI, styling, design.
- Heavy multi-step work: you as orchestrator + Codex CLI executing in a right-hand cmux pane is a solid default setup.
Sending prompts to a TUI agent
- ONE single line — never newlines in the message body. In a TUI, newline = Enter: a multi-line prompt submits at the first line and the rest arrives as fragmented mid-turn steering messages. Use ". " or "; " instead of line breaks, then one explicit enter. For long instructions, write them to a file and send:
read /tmp/task.md and follow it. - Wrap the prompt in plain double quotes — NEVER escaped.
cmux send --surface surface:N "your prompt". The recurring bug is emitting\"— in bash that's literal-broken and dies withunexpected EOF. Inside the prompt, avoid apostrophes and literal double quotes (write "dont", "wont", "lets"); rephrase instead of escaping. If a send failed, the cause was the escaped\", not the quote type. - Exact command names:
cmux send --surface surface:Nthencmux send-key --surface surface:N enter. There is NOsend-surfaceorsend-key-surface.
Polling
Keep sleeps SHORT: start at 3-5s, re-check, repeat. Don't sleep 30. Pi and Hermes (opus-4.8-fast) launch and respond within seconds; scale up only for genuinely heavy tasks. After every check, send the user a one-line status: what the agent is doing and whether it's on track.
Claude Code note: after it finishes, it may prefill a predicted next user message — that draft is Claude, not the user.
Remote VPS
SSH in first and launch the agent ON the VPS (e.g. codex --yolo), then drive that on-box agent. Don't run an agent locally and have it SSH for every step.
The 4 agents (background reference)
All four use the portable SKILL.md standard; project skills win over global.
- Pi (pi.dev, open-source TS): minimal read/write/edit/bash core, self-extends via TS extensions; true BYOK; best-in-class session branch/fork/resume. Skills:
~/.pi/agent/skills/. - Codex CLI (OpenAI, Rust): fastest startup; kernel-level sandboxing;
codex execfor CI; reads AGENTS.md. Skills:~/.codex/skills/. - Claude Code (Anthropic, TS): deepest Claude integration,
.claude/conventions, live skill hot-reload. Skills:~/.claude/skills/. - Hermes (Nous Research, Python): persistent autonomous agent — cross-session memory, built-in scheduler, 40+ tools; can orchestrate the other CLIs as workers. Skills:
~/.hermes/skills/.
Driving interactive CLIs
- Codex, Pi, OpenCode: need
pty=true. - Claude Code: prefer
claude --print --permission-mode bypassPermissions(no PTY).
Example
User request:
Delegate these independent work items with complete context, clear ownership, and progress checks.
Limitations
- Adapted from
davidondrej/skills; verify local paths, tools, credentials, and agent features before acting. - For commands, remote access, scheduling, browser automation, or file-changing workflows, get explicit user approval and confirm the target environment first.
04 同仓库更多 Skills
zero-trust
Implement zero-trust network architecture. Configure identity-based access, micro-segmentation, and continuous verification. Use when implementing modern security architectures.
zcode-delegate
Delegate coding tasks to the Z.AI ZCode CLI only when the user explicitly requests it, while the orchestrator retains review and landing responsibility.
youtube-transcript
Fetch YouTube transcripts through DeepAPI or local fallback tooling and save clean text output.
youtube-transcript-skills
Fetch YouTube video transcripts, search videos/channels, browse channels, and extract playlists via the getyoutubetranscript.com API - free tier, no card required.
youtube-notetaker
Turn YouTube talks into local study notes with slides, transcripts, editable annotations, and a markdown-backed viewer.
youtube-full
Fetch YouTube transcripts, search videos, browse channels, and extract playlists via TranscriptAPI — no yt-dlp, no Google API key, works from any cloud server.