Agent Skills › WingedGuardian/GENesis-AGI › research

research

GitHub

执行深度主题研究，通过界定范围、收集权威源、综合分析与评估可靠性，产出包含关键发现、置信度及行动项的结构化报告，辅助决策或解决复杂问题。

src/genesis/skills/research/SKILL.md WingedGuardian/GENesis-AGI

Trigger Scenarios

用户请求对特定主题进行研究遇到需要理解陌生领域的阻碍计算资源充裕且有待处理的研究任务评估流程标记需深入分析的 WATCH 或 ADOPT 项目

Install

npx skills add WingedGuardian/GENesis-AGI --skill research -g -y

More Options

Non-standard path

npx skills add https://github.com/WingedGuardian/GENesis-AGI/tree/main/src/genesis/skills/research -g -y

Use without installing

npx skills use WingedGuardian/GENesis-AGI@research

指定 Agent (Claude Code)

npx skills add WingedGuardian/GENesis-AGI --skill research -a claude-code -g -y

安装 repo 全部 skill

npx skills add WingedGuardian/GENesis-AGI --all -g -y

预览 repo 内 skill

npx skills add WingedGuardian/GENesis-AGI --list

SKILL.md

Frontmatter

{
    "name": "research",
    "phase": 6,
    "consumer": "cc_background_research",
    "skill_type": "workflow",
    "description": "Deep research on a topic — use when investigating unfamiliar domains, answering complex questions requiring multiple sources, or when an evaluation flags something for deeper analysis"
}

Research

Purpose

Conduct thorough research on a topic, producing a structured summary with sources and actionable takeaways for Genesis.

When to Use

User requests research on a topic.
A blocker requires understanding an unfamiliar domain.
Surplus compute is available and a research task is queued.
An evaluation identified a WATCH or ADOPT item needing deeper analysis.

Workflow

Scope — Define the research question clearly. What specifically do we need to know? What decisions does this inform?
Gather — Search web, documentation, code repositories. Collect primary sources. Prefer official docs over blog posts.
Synthesize — Organize findings into a coherent narrative. Identify consensus vs. conflicting information.
Assess reliability — Note source quality, recency, potential bias.
Extract actionables — What should Genesis do with this information?
Write output — Structured research report.

Output Format

topic: <research question>
date: <YYYY-MM-DD>
summary: <3-5 sentence overview>
key_findings:
  - finding: <finding>
    confidence: high | medium | low
    source: <source reference>
action_items:
  - <concrete next step>
open_questions:
  - <what remains unknown>