Deep Insight — Gerador de Insight Estrutural Profundo
yogsoth-ai/deep-insight
Cinco campanhas de análise profunda para converter gaps em insights estruturados
Ver no GitHubMCP server que transforma gaps de pesquisa em compreensão genuína através de 5 campanhas analíticas (análise de lacunas, geração de insights, análise de limites, análise de sensibilidade, reformulação de problemas). Orquestra 111 skills em camadas de estratégia, tática e SOPs para transformar revisão superficial em entendimento estrutural real.
Instalação
npx convez-ai yogsoth-ai/deep-insight1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.
Métricas
Casos de uso
- Identificar e validar lacunas de pesquisa genuínas vs problemas já resolvidos
- Gerar insights que vão além de revisão de literatura em compreensão real
- Analisar sensibilidade de pressupostos e reformular problemas de pesquisa
- Orquestrar análises complexas com múltiplas estratégias e táticas
Veredito

Sistema sofisticado e modular para pesquisa profunda, mas exige integração com o orquestrador DARE completo para funcionar conforme especificado. Valioso para pesquisadores que precisam transformar gaps em insights estruturados, não apenas listar referências.
README
A tradução para o português deste README ainda está na fila.
🔬 deep-insight
"The important thing in science is not so much to obtain new facts as to discover new ways of thinking about them." — William Lawrence Bragg
Deep Insight is a skill repository for the yogsoth-ai ecosystem that transforms surface-level research gaps into deep structural understanding. It orchestrates 5 analytical campaigns — gap analysis, insight generation, boundary analysis, sensitivity analysis, and problem reformulation — to produce research insights that go beyond literature review into genuine understanding.
🧭 Part of the De-Anthropocentric Research Engine. This repository is one of nine composable research packages that make up DARE — the full autonomous research-orchestration system. DARE bundles this package together with the others into a single self-contained clone, unified under one orchestrator. To use these skills as intended — with the spec-driven orchestrator and cross-package routing — clone the main repository rather than this repo alone.
Architecture
┌─────────────────────────────────────────────────────────────────┐
│ ENTRY.md │
│ (Campaign Router) │
├─────────────────────────────────────────────────────────────────┤
│ │
│ ┌──────────────┐ ┌──────────┐ ┌───────────────────────────┐ │
│ │ gap-analysis │ │ insight │ │ boundary-analysis │ │
│ │ 5 strat │ │ 5 strat │ │ 5 strat │ │
│ │ 3 tactic │ │ 4 tactic │ │ 3 tactic │ │
│ │ 11 SOP │ │ 14 SOP │ │ 12 SOP │ │
│ └──────────────┘ └──────────┘ └───────────────────────────┘ │
│ │
│ ┌──────────────────────────┐ ┌────────────────────────────┐ │
│ │ sensitivity-analysis │ │ problem-reformulation │ │
│ │ 5 strat │ │ 5 strat │ │
│ │ 3 tactic │ │ 3 tactic │ │
│ │ 11 SOP │ │ 10 SOP │ │
│ └──────────────────────────┘ └────────────────────────────┘ │
│ │
│ ┌────────────────────────────────────────────────────────────┐ │
│ │ Shared: 5 import SOPs + 3 shared subagent SOPs │ │
│ └────────────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────────┘
Scale
| Layer | Count |
|---|---|
| Campaigns | 5 |
| Strategies | 25 |
| Tactics | 16 |
| SOPs (subagent) | 58 |
| SOPs (import) | 5 |
| SOPs (shared) | 3 |
| Total skill directories | 111 |
Quick Start
# Clone
git clone https://github.com/yogsoth-ai/deep-insight.git
# Configure MCP servers (copy and fill in API keys)
cp .mcp.example.json .mcp.json
# Use via Claude Code or compatible agent
# Entry point: ENTRY.md routes to campaigns based on research phase
Campaigns
Gap Analysis — Identify, validate, and prioritize research gaps. Distinguishes genuine opportunities from already-solved problems, measures feasibility and impact, ranks by multi-criteria scoring.
Insight — Transform gaps into deep structural understanding. Applies root-cause drilling, stakeholder mapping, tension mining, question reformulation, abstraction laddering, and assumption auditing.
Boundary Analysis — Map the validity envelope of research approaches. Determines where methods work, where they break, and what drives the transition. Uses validity envelope mapping, failure mode analysis, and contextual boundary detection.
Sensitivity Analysis — Quantify which assumptions and parameters matter. Screens parameters, decomposes variance, tests assumption criticality, propagates uncertainty, and identifies decision-sensitive factors.
Problem Reformulation — Question the problem itself. Escapes dominant paradigms, reframes from multiple perspectives, applies dialectical inquiry, assesses wickedness, and discovers appreciative alternatives.
Dependencies
This skill repo imports SOPs from sibling yogsoth-ai repositories:
| Dependency | SOPs Imported |
|---|---|
| web-browsing | web-search, web-research |
| literature-engine | paper-overview, paper-search, paper-research |
| subagent-spawning | spawn-agent (execution runtime) |
| context-management | context-checkpoint (state persistence) |
📄 License
A component of the De-Anthropocentric Research Engine, part of the Yogsoth AI ecosystem. Built by Pthahnix.