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.
npx agentmods add skills/kevinzai/commander/ccc-orchestratenpx skills add KevinZai/commander --skill ccc-orchestrategit clone --depth 1 https://github.com/KevinZai/commanderWrote 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/skills/kevinzai/commander/ccc-orchestrate)<a href="https://agentmods.dev/skills/kevinzai/commander/ccc-orchestrate"><img src="https://agentmods.dev/badge/skills/kevinzai/commander/ccc-orchestrate.svg" alt="Measured on agentmods" 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 | $0.00055 | $0.02389 |
| Opus 5 | $0.00028 | $0.01195 |
| Sonnet 5 | $0.00011 | $0.00478 |
| Haiku 4.5 | $0.00006 | $0.00239 |
Grade A, and why
ccc-orchestrate 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.
How it starts
The opening of the file, as written. The whole thing — 179 lines — stays where its author put it; the contents beside it link to each section on GitHub.
$ccc-orchestrate — Cross-Runtime Orchestrator/Executor
Orchestrate splits work across two different model runtimes, not just two tiers of the same one. A capable model (Fable 5 or Opus 5 — the calling session, i.e. "you") compresses a fuzzy request into a tight, testable goal file. Execution then happens outside this runtime entirely — on GPT-5.6 Sol via the codex CLI, or on a Sonnet subagent if codex isn't available. The orchestrator never writes production code itself.
CC Commander · Cross-Runtime Orchestrate · Docs
This is
$ccc-plan-execgeneralized past Claude-only: instead of plan (cheap Claude) → execute (capable Claude), it's plan (Fable/Opus) → execute (a different runtime — GPT-5.6 or Sonnet). The goal file is the interface between the two systems, so it has to have real teeth.
Why cross-runtime
Fable/Opus is expensive per token but excellent at compressing a fuzzy request into a tight, unambiguous spec. GPT-5.6 Sol at high/xhigh reasoning is a strong, comparatively cheaper implementer for well-specified work. The split only pays for itself if the goal file is genuinely load-bearing — vague or hand-wavy acceptance criteria collapse the whole advantage, because the orchestrator ends up re-explaining the task on every re-dispatch anyway. Write the goal file as if the executor has zero conversational context, because on the codex path it does.
Model guidance (GPT-5.6 family, commander/adapters/codex/models.js): default the executor to Sol for real implementation work (deep dev — flagship reasoning). Drop to Terra for balanced, well-specified tasks where Sol's extra reasoning wouldn't change the output. Reserve Luna for cheap, mechanical steps (renames, boilerplate, formatting) inside a larger loop — never for the primary executor role.
The two roles
- Orchestrator (Fable 5, effort high, or Opus 5 — this session) — reads the request, asks clarifying questions if the scope is fuzzy, and writes a structured goal file to disk. Never edits production code in this mode.
- Executor (GPT-5.6 Sol via codex, or Sonnet subagent) — reads the goal file verbatim as its entire brief and implements it. Does not re-plan, re-scope, or renegotiate the acceptance criteria.
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.
- 4d ago First seen · 179 lines · 55 tokens per session scan A aee7b01a803e
ccc-orchestrate is a skill published in the GitHub repository KevinZai/commander (6 stars, last pushed today), licensed MIT. It adds 55 tokens to every session and 2,389 once invoked, about $0.0003 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.
Other skills, from other repositories
claude-code-session-broker
Use when running Arcgentic V2 in Claude Code and fixed Planner, Developer, and Auditor role sessions must be coordinated through a broker.
verify-gates
Runs the mechanical quality gates that the arcgentic state machine requires for state transitions. Invoked indirectly by transition.sh OR directly by orchestrator agent before declaring a state transition. Use when about to call transition.sh OR when manually verifying that a round artifact meets the gate criteria.…
agency-roster
Use when a round references agency-agents catalogs, role-family routing, multi-agent identity prompts, or English/Chinese specialist role catalogs.
cross-session-handoff
Read, write, snapshot, and lock .arcgentic/state.yaml across planner, dev, audit, and optional test sessions.
hook-template
Generate hook script from template. Use when adding a new hook, wiring a PreToolUse/PostToolUse/Stop/Notification hook, or scaffolding hook config for settings.json.
check-environment
Verify Claude, Codex, and Grok availability plus Director guidance, relay, agents, and skills. Audit optional hooks only when selected. Use after installation or when a native surface misbehaves.