skills.rest
ginaecho
ginaechoCOMMUNITY
Explore all skills by this author →
GitHub

fan-out-research

Splits research questions across dynamically spawned helper subagents and delivers a consolidated report.

1|Updated Jul 1, 2026
Search Tags:#research#subagents#multi-agent#report#planner#fan-out#dynamic spawning
One-click install
npx skills add https://github.com/ginaecho/skill-achievability-compiler --skill fan-out-research-ginaecho
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: fan-out-research
Source: https://github.com/ginaecho/skill-achievability-compiler/tree/main/corpus/skills/spawn_helpers
Command: npx skills add https://github.com/ginaecho/skill-achievability-compiler --skill fan-out-research-ginaecho

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Research questions that are too broad for a single agent to handle efficiently need to be divided among multiple workers, but the number of sub-questions often cannot be known in advance. This Skill coordinates a planner agent that decomposes a research question at run time, spawns one helper subagent per part, and delivers a final report. ## Core Features & Use Cases - Dynamic Subagent Spawning: Creates fresh helper subagents during execution, with no fixed or bounded number decided beforehand. - Planner-Helper Workflow: A planner breaks the question into parts, assigns each part to a helper, collects the results, and marks the report delivered via the deliver tool. - Use Case: Given a broad question like "compare the pricing, features, and compliance posture of ten cloud vendors", the planner spawns one helper per vendor, gathers their findings, and delivers a single consolidated report. ## Quick Start Research this topic by fanning it out to helper subagents for each sub-question, then deliver a consolidated report.

Frequently Asked Questions about fan-out-research

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I split a research task across multiple subagents?▼

Use a planner agent that breaks the research question into parts at run time and spawns one fresh helper subagent per part. The planner then collects each helper's results and delivers a consolidated report.

How many subagents does this research workflow create?▼

The number is not fixed in advance. The planner decides how many parts the question has during execution and spawns one helper per part, with no upper bound on how many it may create.

What tools does the fan-out research workflow require?▼

The workflow requires a deliver tool that marks the final report as delivered, plus a runtime capable of spawning helper subagents at run time. Without subagent-spawning capability in the environment, the skill cannot execute as written.

When should I not use dynamic subagent fan-out for research?▼

Avoid it for narrow questions answerable by a single agent, since spawning helpers adds coordination overhead. It also does not fit environments that lack subagent-spawning support or that require a fixed, predeclared set of participants.

Related Skills

Explore More Education & Research Skills→
TrecekTrecek
1

make-req

Decomposes task descriptions into structured, independently plannable requirement groups.

Community
Advanced
dralkhdralkh
74

markitdown

Convert documents, media, and archives into clean Markdown.

Community
Advanced
mronkkomronkko
20

fact-check

Verify manuscript citations against Zotero/MCP sources and audit quantitative claims.

Community
Advanced
tamireslaeao-progtamireslaeao-prog

prose

Orchestrate multi-agent AI workflows from a single .prose program.

Community
Advanced