nested-orchestration

A workflow for giving one parent agent permission to start and supervise its own child agents. The parent can monitor those agents, read their results, and view the structure of its agent group.

In plain words
What is it for?
Use it to build an agent hierarchy in which a parent launches multiple sub-agents, gathers their outputs, monitors them, and inspects the resulting session tree.
Why use it?
It supports tasks that need several layers of delegation or continued supervision. The parent can manage the child work without exposing broader shell, file, or remote-control access.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/agentproto/ts/nested-orchestration
Any agent
npx skills add agentproto/ts --skill nested-orchestration
Clone the repo
git clone --depth 1 https://github.com/agentproto/ts

Made for: Claude Code, Codex.

Per session 162 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,207 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00162 $0.02207
Opus 5 $0.00081 $0.01104
Sonnet 5 $0.00032 $0.00441
Haiku 4.5 $0.00016 $0.00221

Measured 2d ago against content hash 223624e47bd2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

nested-orchestration scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

packages/skill-pack-agentproto/src/skills/nested-orchestration/SKILL.md · 183 lines

How it starts

The opening of the file, as written. The whole thing — 183 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Nested orchestration (orchestrator-of-orchestrators)

Methodology + commands for turning an agent into a scoped orchestrator: it spawns its own sub-agents, supervises them (session_monitor), reads their outputs, and sees its subtree (session_tree). Distilled from a real session where every case below was proven live.

To be distinguished from the agent-session-orchestration-agentproto skill (flat orchestration: it's you, in cowork, driving the agents). Here we add a tier: you delegate the orchestration to a parent agent, which drives children. Useful when the breakdown is deep, when you want to offload the polling from your own context, or for a workflow that must run without you at every turn.

The principle in one line

parent claude-code (orchestrator:true) → spawns N children → session_monitor (fan-in) → reads the outputs → session_tree

The daemon mints a scope-token per orchestrator-child, injects the URL of a scoped sub-gateway into the parent's session (alongside any mcpServers you pass), and revokes the token on exit. The parent only receives a curated subset of orchestration tools — never shell / fs / remote / import / terminal.

Golden rule — the parent MUST be claude-code

Proven broken with hermes, working with claude-code. A hermes parent ignores the mcpServers field injected over ACP: it sees its own tools but not the orchestration gateway → it cannot spawn a sub-agent. claude-code mounts the gateway correctly (the ACP fix "mcpServers wire shape for session/new" was on the claude-code side). So: nesting ⇒ parent = claude-code. For the child, any adapter will do (cheap haiku for trivial work, hermes/lightweight for code — see light-coder-orchestration).

Making a parent an orchestrator

agent_start({
  adapter: "claude-code",
  model:   "claude-sonnet-4-6",   // reliable parent for driving
  orchestrator: true,             // ← auto-mounts the scoped sub-gateway
  cwd:     "<absolute HOST path>",
  label:   "parent-…",
  prompt:  "<orchestration brief>"
})

Read the full file on GitHub · 183 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 183 lines · 162 tokens per session scan A 223624e47bd2

Subscribe to this mod's changes

nested-orchestration is a skill published in the GitHub repository agentproto/ts (5 stars, last pushed 2d ago), licensed Apache-2.0. It adds 162 tokens to every session and 2,207 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

strands-review

Local preview of the strands-agents/devtools /strands review agent. Body is the upstream Task Reviewer SOP verbatim — do not paraphrase. Use when the user types /strands-review, asks for a "strands review" of a PR, or wants to anticipate what the remote /strands review GitHub Action will flag. Findings are close but…

strands-agents/harness-sdk · 125 tokens

docs-writer

Draft or rewrite Strands Agents documentation pages. Use when writing new doc pages, rewriting pages that failed audit, drafting sections for existing pages, or writing blog posts and release notes about Strands. Also triggers on "write a doc", "draft a page", "rewrite the quickstart", "add a tutorial for X"…

strands-agents/harness-sdk · 74 tokens

pr-create

Creates a GitHub pull request using the gh CLI. Use when the user asks to create, open, or submit a PR on GitHub.

strands-agents/harness-sdk · 32 tokens

document-service

This skill should be used when the user asks to "analyze this codebase", "document this service", "generate technical docs", "I inherited this code", "help me understand this system", "create docs for this project", "what does this system look like", "onboard me to this codebase", "this codebase has no docs"…

awslabs/agent-plugins · 141 tokens

aws-step-functions

Build workflows with AWS Step Functions state machines using the JSONata query language. Covers Amazon States Language (ASL) structure, state types, variables, data transformation, error handling, AWS service integration, and migrating from the JSONPath to the JSONata query language.

awslabs/agent-plugins · 57 tokens

hyperpod-version-checker

Check and compare software component versions on SageMaker HyperPod cluster nodes - NVIDIA drivers, CUDA toolkit, cuDNN, NCCL, EFA, AWS OFI NCCL, GDRCopy, MPI, Neuron SDK (Trainium/Inferentia), Python, and PyTorch. Use when checking component versions, verifying CUDA/driver compatibility, detecting version mismatches…

awslabs/agent-plugins · 120 tokens