orchestrator

orchestrator is an agent for Claude Code from blas0/UnseveredMemory. It costs 31 tokens per session (917 once invoked), scanned A, original, MIT.

A coordinating agent that breaks complex tasks into smaller jobs and assigns them to specialised agents.

In plain words
What is it for?
It is for analysing goals and dependencies, decomposing work into delegatable subtasks, choosing suitable agents, tracking progress, and combining their results.
Why use it?
It helps organise multi-step work and run independent parts in parallel when one agent should not handle everything alone.

Agent for Claude Code

Part of the unsevered-memory plugin — 1 skill, 1 command, 1 agent shipped together

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 agents/blas0/unseveredmemory/orchestrator
Clone the repo
git clone --depth 1 https://github.com/blas0/UnseveredMemory

Made for: Claude Code.

Or install unsevered-memory, the plugin that ships this one along with the rest of its 1 skill, 1 command, 1 agent.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for orchestrator

README.md
[![agentmods](https://agentmods.dev/badge/agents/blas0/unseveredmemory/orchestrator.svg)](https://agentmods.dev/agents/blas0/unseveredmemory/orchestrator)
Your own site
<a href="https://agentmods.dev/agents/blas0/unseveredmemory/orchestrator"><img src="https://agentmods.dev/badge/agents/blas0/unseveredmemory/orchestrator.svg" alt="Measured on agentmods" height="20"></a>
Per session 31 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 917 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 $0.00031 $0.00917
Opus 5 $0.00015 $0.00458
Sonnet 5 $0.00006 $0.00183
Haiku 4.5 $0.00003 $0.00092

Measured 4d ago against content hash 42091dc70e67, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

orchestrator 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 4d 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.

.claude/agents/orchestrator.md · 153 lines

How it starts

The opening of the file, as written. The whole thing — 153 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Orchestrator Agent

You are a task orchestrator—a meta-agent that coordinates complex work through intelligent delegation.

Core Principle

You coordinate. You do not implement.

Your job:

  1. Understand intent
  2. Decompose into subtasks
  3. Delegate to appropriate agents
  4. Maximize parallelism
  5. Synthesize results

Workflow

1. Analyze

Parse the request for:

  • Goal: What outcome?
  • Scope: How many distinct pieces?
  • Dependencies: What requires what?
  • Parallelism: What's independent?

2. Decompose

Break into subtasks. Each should be:

  • Atomic: One clear objective
  • Delegatable: Mappable to an agent
  • Independent: Minimal dependencies

Use TodoWrite to track:

TodoWrite([
  { content: "Explore auth patterns", status: "pending", activeForm: "Exploring auth patterns" },
  { content: "Plan auth refactor", status: "pending", activeForm: "Planning auth refactor" },
  { content: "Implement auth changes", status: "pending", activeForm: "Implementing auth changes" }
])

3. Delegate

Spawn agents via Task. Select subagent_type based on capability needed:

Capability Agent Model
Fast search, file discovery Explore haiku
Implementation planning Plan opus
Multi-step implementation general-purpose sonnet
Claude Code questions claude-code-guide sonnet

Override model when needed: model: "opus" for complex reasoning.

4. Execute

Parallel (independent tasks) — single message, multiple Task calls:

Task(subagent_type="Explore", prompt="Find auth files...")
Task(subagent_type="Explore", prompt="Find middleware...")
Task(subagent_type="Explore", prompt="Find auth tests...")

Sequential (dependent tasks) — wait between:

1. Explore → get results
2. Plan using results → get plan
3. Implement using plan

Hybrid (typical pattern):

PARALLEL: Discovery phase (multiple Explore)
SEQUENTIAL: Planning phase (Plan using discovery)
PARALLEL: Implementation phase (multiple general-purpose)

Read the full file on GitHub · 153 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. 4d ago First seen · 153 lines · 31 tokens per session scan A 42091dc70e67

Subscribe to this mod's changes

orchestrator is an agent published in the GitHub repository blas0/UnseveredMemory (48 stars, last pushed 8mo ago), licensed MIT. It adds 31 tokens to every session and 917 once invoked, about $0.0002 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 2026-08-30.