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 skills add dryvist/claude-code-plugins --skill premium-agent-orchestrationgit clone --depth 1 https://github.com/dryvist/claude-code-pluginsWrote 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/dryvist/claude-code-plugins/premium-agent-orchestration)<a href="https://agentmods.dev/skills/dryvist/claude-code-plugins/premium-agent-orchestration"><img src="https://agentmods.dev/badge/skills/dryvist/claude-code-plugins/premium-agent-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/skills/dryvist/claude-code-plugins/premium-agent-orchestration"><img src="https://agentmods.dev/badge/skills/dryvist/claude-code-plugins/premium-agent-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 | $0.00083 | $0.01272 |
| Opus 5.5 | $0.00033 | $0.00509 |
| Sonnet 5.5 | $0.00017 | $0.00254 |
| Haiku 4.5 | $0.00008 | $0.00127 |
Grade A, and why
premium-agent-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 today.
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 — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Premium Agent Orchestration
Treat the model running the current session as the senior decision-maker,
whatever it is — model- and vendor-agnostic. This skill is the
orchestrator case only: independent work that together exceeds one
context window, or a marathon multi-workstream coordination session.
Not this skill: serial work with a few
hard decisions, or one chain that fits a single context — tune effort on
one model instead. The baseline to beat is always the frontier model at
high effort alone; measure against that first.
Economics
- Pays when combined input exceeds one context window.
- Measured gains: 47–55% cost cut on large-corpus work; on routine batches, half the average cost and a third of the worst-case (tail) cost versus one agent working serially.
- Measured cost: 10–12 accuracy points when workers are weaker than the lead. Mitigate by keeping merging and final judgment on the lead — never let a worker's output ship unreviewed.
Senior Model Owns
The flagship keeps judgment on intent, design, tradeoffs, risk, and final approval. The lead checks every result against its yes/no check and sends failures back with the named shortfall; a delegate's claim isn't evidence.
Delegate strictly downward — never spawn a peer at your own tier. Send quick lookups, exploration, and search to the lowest capable executor, not just one down — that work costs volume, not thought. Forks run on the lead's model. Prefer the router for checkable, low-skill work; discover availability live, never hard-code a physical model ID or a static task-to-model table.
Concurrency and batching
- Parallelize ready work. Launch all independent, ready work in parallel; sequence only for real dependencies or same-file conflicts. Group related lookups within a worker when they form one task. Probe the spawn substrate before the first fan-out; if spawning fails, take the solo path. Never cap the number of workers. Gate a large fan-out only on measured machine load (CPU and memory under 80% of capacity); spawn failures and quota or rate limits are the other signals (see Substrate Resilience below). Batch many small related tasks into one routed Agent call as a task list.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- today Changed · -14 lines a97e0b3b59f7
- 5d ago Changed · +2 lines 50c88e34c589
- 10d ago Changed · -34 lines 6e181df12ec1
- 11d ago Changed · +3 lines 496245c401b2
- 13d ago Changed · +4 lines · +1 tokens per session a9302920127b
- 15d ago Changed · -17 lines · -5 tokens per session 3631374d9901
- 18d ago Changed · +10 lines 3e6193788eaf
- 1mo ago Changed · -119 lines · +21 tokens per session 0a6ae7c5cdcf
- 1mo ago Changed · +14 lines acb537de8d40
- 1mo ago First seen · 265 lines · 66 tokens per session scan A 99c480d987b2
premium-agent-orchestration is a skill published in the GitHub repository dryvist/claude-code-plugins (3 stars, last pushed today), licensed Apache-2.0. It adds 83 tokens to every session and 1,272 once invoked, about $0.0003 per session on Opus 5.5. 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
detect-github-url
Detects when the user pastes or mentions a GitHub URL that may contain Claude skills or plugins. Triggers on messages containing github.com URLs, especially when user says things like "here's a repo", "check out this repo", "load this", "use these skills", "I found these skills", or simply pastes a github.com link.…
example-feature-generator
Generate boilerplate for new features following project conventions. Use when user says "create feature", "add feature", or "new feature scaffolding".
media-gemini-3-image-prompt
A prompt builder for Gemini 3 Pro Image, also known as Nano Banana Pro, Google's image-generation model. It expands a short request into five parts covering the subject, action, setting, composition, lighting, style, and text constraints.
commerce-jtbd-persona
An e-commerce customer-analysis skill that studies reviews and questions to identify why people buy and to create concrete customer profiles. JTBD means “jobs to be done”: the practical goal a customer is trying to accomplish.
data-realestate
A lookup tool for reported South Korean property transactions using Ministry of Land data. It covers sales and rental deals for apartments, officetels, multi-family homes, houses, and commercial or office properties.
design-sync-upload
An uploader for design-system files such as DESIGN.md, tokens, logos, fonts, and images into Claude Design. It can either use an authenticated connection or prepare a folder and guide for manual upload.