Agent Orchestrator is a local desktop workspace for running and supervising multiple coding agents, giving each task its own session, workspace, branch, and feedback loop. It is for developers coordinating agent fleets across repositories, pull requests, CI runs, reviews, and merges. The catalogue entries are commands, skills, and instructions for working with this orchestrator.
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.
git clone --depth 1 https://github.com/Untrivial-ai/agent-orchestratorWrote 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.
[](https://agentmods.dev/commands/untrivial-ai/agent-orchestrator/orchestration)<a href="https://agentmods.dev/commands/untrivial-ai/agent-orchestrator/orchestration"><img src="https://agentmods.dev/badge/commands/untrivial-ai/agent-orchestrator/orchestration/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/untrivial-ai/agent-orchestrator/orchestration"><img src="https://agentmods.dev/badge/commands/untrivial-ai/agent-orchestrator/orchestration.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00000 | $0.00993 |
| Opus 5.5 | $0.00000 | $0.00397 |
| Sonnet 5.5 | $0.00000 | $0.00199 |
| Haiku 4.5 | $0.00000 | $0.00099 |
Grade A, and why
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 11d 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.
How it starts
The opening of the file, as written. The whole thing — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Orchestration: spawn, list, send, report, kill
All five commands talk to the AO control plane with this session's rotating
worker credential. spawn, list, send, and kill require the
worker:orchestrate scope, which only orchestrator sessions hold; report
requires a parent orchestrator. A command run without the required scope fails
with SCOPE_REQUIRED.
ao spawn — create a worker session
ao spawn --name "<label>" --prompt "<clear worker task>" [--agent <harness>] [--mode standard|trusted]
--name(required): the label the human sees in the sidebar. Keep it short and specific. Twenty characters or fewer is ideal; 100 characters is the hard cap. Count it before running the command; an over-long label wastes a turn on a rejection.--prompt(required): the worker's complete task. Write it like a brief for a competent engineer with no other context: goal, constraints, and the expected outcome (usually a pull request). Before you name a specific file, class, or path, confirm it exists in the checked-out repository (ls,find,grep). A prompt pointing at a file that does not exist wastes the whole worker.--agent/--harness:claude-code,codex,cursor, oropencode. If omitted, the project-configured worker harness is used; when the project has no worker harness, the control plane falls back toclaude-code.--mode:trusted(default) orstandard.
Prints spawned <session-id> (<status>). The worker provisions its own
sandbox, checks out the project repository on a fresh branch, and starts on the
prompt — typically live within ~10 seconds. You do not need to wait: a send
issued before it is ready is queued and delivered on startup.
Errors: agent_provider_required means the org has no validated credential for
that harness — pick a different --agent or report the blocker to the human.
SANDBOX_QUOTA_EXCEEDED means the org hit its sandbox cap — kill finished
workers first.
ao list — inspect your workers
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.
- 11d ago Changed · +5 lines ea9297495f3e
- 28d ago Changed · +6 lines fee912eb9a49
- 1mo ago First seen · 85 lines · 0 tokens per session scan A 4b564cc8774d
orchestration is a command published in the GitHub repository Untrivial-ai/agent-orchestrator (13,060 stars, last pushed 5d ago), licensed Apache-2.0. It costs nothing until one of its globs matches a file; then it loads 993 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 2026-09-14.
Other commands, from other repositories
sync-all
Sync skills, commands, MCP servers, and preferences between Claude, Codex, Copilot, and Cursor in one operation.
sync-to-cursor
Sync skills, commands, agents, hooks, rules, and MCP from Claude, Codex, or Copilot to Cursor IDE.
skill-diff
Compare a skill across Claude, Codex, Copilot, and Cursor to show differences.
skill-trace-status
Check skill loading status and instrumentation across CLI environments.
sync-from-copilot
Sync skills and instructions from GitHub Copilot CLI to Claude, Codex, or Cursor.
add-tests
Add tests using a structured testing approach.