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 ellmos-ai/build-your-users-mind --skill swarm-operationsgit clone --depth 1 https://github.com/ellmos-ai/build-your-users-mindWrote 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/ellmos-ai/build-your-users-mind/swarm-operations)<a href="https://agentmods.dev/skills/ellmos-ai/build-your-users-mind/swarm-operations"><img src="https://agentmods.dev/badge/skills/ellmos-ai/build-your-users-mind/swarm-operations/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/ellmos-ai/build-your-users-mind/swarm-operations"><img src="https://agentmods.dev/badge/skills/ellmos-ai/build-your-users-mind/swarm-operations.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 | $0.00064 | $0.01404 |
| Opus 5 | $0.00032 | $0.00702 |
| Sonnet 5 | $0.00013 | $0.00281 |
| Haiku 4.5 | $0.00006 | $0.00140 |
Grade A, and why
swarm-operations 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 10d 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 — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Swarm Operations: 5 patterns for multi-agent systems
Coordination strategies for AI agents working in parallel or hierarchically — from simple task distribution to emergent swarm patterns.
Bundled dependency of build-your-users-mind. The module's stage-2 classification uses pattern 2 (Hierarchy) + pattern 3 (Stigmergy): domain leads direct chunk workers, and agents coordinate through shared files rather than direct messaging.
Overview of the 5 patterns
| # | Pattern | Agents | Coordination | Best for |
|---|---|---|---|---|
| 1 | Epstein method | 3-5 | Central (orchestrator) | Knowledge building, research |
| 2 | Hierarchy swarm | 5-20 | Tree structure | Large projects, refactoring |
| 3 | Stigmergy | 3-10 | Decentralized (environment) | Exploration, data crawling |
| 4 | Consensus | 3-7 | Peer-to-peer | Decisions, review |
| 5 | Specialist swarm | 3-8 | Hub-and-spoke | Cross-domain |
Pattern 1: Epstein method
Principle: one orchestrator distributes subtasks to specialized workers; results are collected, deduplicated and merged into a whole. Roles: Orchestrator (split, distribute, synthesize) · Worker (handle subtask, return structured result) · Validator (optional, quality/consistency check). Use when: research with several independent aspects, knowledge building, multi-source extraction. Limits: orchestrator is a single point of failure; overhead below 3 subtasks; dependent subtasks hard to parallelize.
Pattern 2: Hierarchy swarm
Principle: tree of boss nodes coordinating sub-teams; each boss knows only its direct reports. Levels: Project boss (overall plan, final result) · Team lead (coordinate a subarea) · Worker (single task). Implementation: decompose into 2-4 areas → one lead per area → leads assign workers → aggregate bottom-up (worker → lead → boss). Use when: large projects (>10 files), refactoring across modules, documenting whole systems, natural hierarchy. Limits: cross-level communication overhead; the big picture can be lost at lower levels; needs ≥5 agents to pay off.
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.
- 10d ago First seen · 99 lines · 64 tokens per session scan A 55b994ee0780
swarm-operations is a skill published in the GitHub repository ellmos-ai/build-your-users-mind (3 stars, last pushed 4d ago), licensed MIT. It adds 64 tokens to every session and 1,404 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
trulens-evaluation-setup
Configure feedback functions and selectors for TruLens evaluations.
self-improver
Review affiliate campaign results and improve strategy. Triggers on: "review my results", "what went wrong", "how to improve conversions", "analyze my campaign", "affiliate retrospective", "why am I not converting", "improve my strategy", "what should I change", "campaign review", "optimize my approach", "learn from…
md-review
Run the standard markdown review loop end-to-end with the mdxg-redline CLI: send a markdown document to a human reviewer, wait (via Bash polling) for the feedback.json they export, then apply their location-aware inline comments back into the markdown and revise it. Always use this skill when the user wants…
decision-avatar
Bei vorhandenem und ausdrücklich autorisiertem lokalem Entscheidungsprofil: wiederkehrende Entscheidungen aus belegtem Feedback vorhersagen, Konfidenz kalibrieren und Vorhersage, Entscheidung und Ausführung strikt trennen.
build-your-users-mind
Verweist auf das öffentliche, anbieterneutrale Modul build-your-users-mind: ein datenschutzbewusstes Verfahren zum Aufbau eines empirischen Theory-of-Mind-Präferenzmodells eines autorisierten Nutzers aus dessen eigenen Interaktionsprotokollen. Verwenden, wenn eine zuständige Person ein Entscheidungsavatar-Profil…
ai-feedback-design-principles
Audit and redesign AI-generated feedback for pedagogical quality, timing, and learning impact. Use when building or reviewing automated feedback in digital learning tools.