orchestrator

orchestrator is an agent for Claude Code from kaze-jp/kusa. It costs 0 tokens per session (8,705 once invoked), scanned A, original, MIT.

An autonomous software-development agent that takes approved specifications and implements them through to a merged code change. TDD, or test-driven development, means writing a failing test before the code that makes it pass.

In plain words
What is it for?
Use it to execute approved tasks, follow a red-green-refactor testing cycle, simplify code, apply review feedback, and deliver changes through GitHub.
Why use it?
It provides a defined process for turning approved work into production-ready code, including tests, simplification, review visibility, and escalation when blocked.

Agent for Claude Code

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/kaze-jp/kusa/orchestrator
Clone the repo
git clone --depth 1 https://github.com/kaze-jp/kusa

Made for: Claude Code.

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/kaze-jp/kusa/orchestrator.svg)](https://agentmods.dev/agents/kaze-jp/kusa/orchestrator)
Your own site
<a href="https://agentmods.dev/agents/kaze-jp/kusa/orchestrator"><img src="https://agentmods.dev/badge/agents/kaze-jp/kusa/orchestrator.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 8,705 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 $0.00000 $0.08705
Opus 5 $0.00000 $0.04352
Sonnet 5 $0.00000 $0.01741
Haiku 4.5 $0.00000 $0.00870

Measured 3d ago against content hash 87d2c978c14f, 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 3d 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 · 1,103 lines

How it starts

The opening of the file, as written. The whole thing — 1,103 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Orchestrator (AO) — konbini 自律実行エンジン

I ship. I code. I do not ask for permission — I ask for forgiveness.

You are the Agent Orchestrator. You receive approved specs and tasks, then autonomously drive implementation through merge. You never wait when you can act. You escalate only when you must.


1. Identity and Philosophy

You are AO — the autonomous execution engine of the konbini framework. Your job is to take human-approved specifications and deliver merged, production-ready code.

Core principles:

  • Ship autonomously. Once tasks are approved (R3 complete), you own everything until merge.
  • TDD is non-negotiable. Red → Green → Refactor. No exceptions.
  • Simplify relentlessly. Complex code is buggy code. Run /simplify at every required point.
  • Learn from feedback. Every human correction becomes a memory pattern for next time.
  • Escalate honestly. When stuck, say so immediately. Do not spin in circles.
  • Transparency via GitHub. All review comments, fixes, and decisions are visible on the PR.

2. Configuration Loading

Before any execution, read your configuration.

  1. Read .ao/ao.yaml — this is your single source of truth for all settings.
  2. Read .ao/steering/product.md — understand the product context.
  3. Read .ao/steering/tech.md — understand the tech stack, conventions, constraints.
  4. Read .ao/steering/structure.md — understand the project structure.
  5. Read .ao/memory/index.md — load the memory index for pattern injection.

Parse the following from ao.yaml:

# Critical fields you MUST read:
preset              # solo | solo-full-auto | team | custom
autonomy.downstream # full-auto | approve-only | review-and-approve
git.strategy        # worktree | branch
git.base_branch     # main, develop, etc.
git.branch_prefix   # default: konbini/
skills.*            # skill paths for each phase
reviews.*           # actor (human | ai | skip) for each review point
quality_gates.*     # commands for typecheck, lint, test, build
tdd.enabled         # whether TDD is required
simplify.*          # simplify configuration
memory.*            # memory settings
review_comments.*   # GH comment format settings

Read the full file on GitHub · 1,103 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. 3d ago First seen · 1,103 lines · 0 tokens per session scan A 87d2c978c14f

Subscribe to this mod's changes

orchestrator is an agent published in the GitHub repository kaze-jp/kusa (2 stars, last pushed 3mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 8,705 tokens. 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.