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 neuromechanist/research-skills --skill agent-fanoutgit clone --depth 1 https://github.com/neuromechanist/research-skillsWrote 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/neuromechanist/research-skills/agent-fanout)<a href="https://agentmods.dev/skills/neuromechanist/research-skills/agent-fanout"><img src="https://agentmods.dev/badge/skills/neuromechanist/research-skills/agent-fanout/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/neuromechanist/research-skills/agent-fanout"><img src="https://agentmods.dev/badge/skills/neuromechanist/research-skills/agent-fanout.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Privilege Escalation · line 134 Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.Fix: Remove references to credential paths. Use environment variables or secrets managers. For docs, use placeholder paths (e.g., /path/to/config). Never load .env or token files in production code paths.
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 | $0.00113 | $0.02889 |
| Opus 5 | $0.00056 | $0.01444 |
| Sonnet 5 | $0.00023 | $0.00578 |
| Haiku 4.5 | $0.00011 | $0.00289 |
Grade A, and why
agent-fanout 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 12d 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 — 210 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Fan-Out and Teams
Orchestrate multiple subagents or teammates to explore, implement, review, and validate in parallel. This skill makes every judgment call explicit: when to fan out, how many agents, which model, what to put in each prompt, how to supervise, and how to combine results.
When to fan out (decision table)
| Situation | Action |
|---|---|
| Single bounded question ("where is X defined?") | No fan-out. Search directly or use one read-only explorer. |
| Open-ended audit/review of a system with 2+ independent subsystems | One read-only explorer per subsystem, in parallel. |
| Multiple root-caused issues, each fixable independently | One implementer per issue, each in its own git worktree. |
| One pull request (PR) ready for review | One review panel (2-5 reviewers, see Review panels below). |
| Sequential work where step N needs step N-1's output | Do NOT parallelize. Run one agent at a time or do it inline. |
| Task needs secrets, deploy rights, or user-only credentials | Do NOT delegate. Keep it in the main session (mark the task "owner: lead"). |
Do not fan out for work you can finish inline in a few minutes; a subagent costs setup, supervision, and synthesis time.
Hard limits (compute BEFORE launching)
- Before any fan-out, compute the worst case:
finders x max findings per finder x verifiers per finding + implementers + reviewers. Write the number down in your plan. - Budget: 10-20 agents per run for routine work; hard cap 40. If the
math exceeds the routine budget, say so and justify it; if it exceeds 40,
cut scope before launching: fewer lenses, findings capped per agent
(
maxItems-style limits in the prompt), one verification vote instead of three. Going past 40 requires explicit user approval in the same conversation. If the user's own configuration states a stricter cap, the stricter number wins. - Prefer waves of 10 or fewer concurrent agents. Finish and synthesize a wave before launching the next.
- Every phase plan states its own agent budget and routing up front (for example "Sol lead + Terra phase planner + Luna implementer/reviewer, ~4 agents total").
- If you hit a rate limit: stop spawning, schedule one backoff wait, resume staggered. Narrate as a status update, not a question.
What ships with it
2 files 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.
- 12d ago First seen · 210 lines · 113 tokens per session scan A e17114b3d3f7
agent-fanout is a skill published in the GitHub repository neuromechanist/research-skills (45 stars, last pushed 9d ago), licensed BSD-3-Clause. It adds 113 tokens to every session and 2,889 once invoked, about $0.0006 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 2026-08-30.
Other skills, from other repositories
n8n-architect
Use when the user explicitly wants to create, edit, validate, sync, or troubleshoot n8n workflows, asks about n8n nodes or automation, or wants to use n8n-as-code in the current context root.
ai-navigation
Use when implementing AI movement — NavigationAgent2D/3D, steering behaviors, behavior trees, and patrol patterns.
phantom-camera
Use when using the Phantom Camera addon — PhantomCamera2D/3D with priority-based switching, follow and look-at modes, and tween transitions.
3d-essentials
Use when working with 3D-specific systems — materials, lighting, shadows, environment, global illumination, fog, LOD, occlusion culling, and decals in Godot 4.3+.
ability-system
Use when building character abilities — Resource-based abilities with cost/cooldown/cast, buffs/debuffs, stat modifiers, gameplay tags, and HUD binding.
addon-development
Use when creating Godot editor plugins — EditorPlugin, @tool scripts, custom inspectors, and dock panels.