convez.ai
MCP ServerFinanceiroAvançado
gmanish10

Equity Research Skill — Análise de Ações Agressiva

gmanish10/equity-research-skill

Análise de equity research opinionada para carteiras agressivas com risco explícito

Ver no GitHub
InicialPluginScore · de 0 a 100

MCP server Python que oferece análise fundamentada de ações, ETFs, opções e portfólios através de uma metodologia de 7 fases (negócio, setor, fundamentals, técnica, opções, sentimento, veredito). Integra dados do Yahoo Finance com web search para pesquisa de risco e oportunidade em carteiras agressivas de crescimento, sempre com callouts de risco quantificados.

Instalação

$npx convez-ai gmanish10/equity-research-skill

1 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

Estrelas
0
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 5 meses
Versão
v1.0.0
Issues abertas
0
Contribuidores
2

Casos de uso

  • Análise fundamentada de ações com rating multi-horizonte e risco quantificado
  • Avaliação de portfólios (correlação, concentração, beta, drawdown, fatores)
  • Geração de ideias de investimento via screens de mercado com workup completo
  • Análise de opções (IV rank, skew, fluxo, max pain, movimentos implícitos)
  • Revisão de ETFs e fit estratégico na carteira existente
  • Planos de rebalanceamento com gatilhos, prazos e callouts de risco por linha

Veredito

Ferramenta robusta e opinada para quem busca análise de equity research em Python com metodologia estruturada e risco explícito. Requer setup técnico (Yahoo Finance MCP, deps Python), disciplina para interpretação (não é aconselhamento personalizado) e perfil de risco agressivo. Disclaimer claro: dados com ~15 min de delay, recomendações educacionais apenas.

README

A tradução para o português deste README ainda está na fila.

Equity Research Plugin (Aggressive-Growth Edition)

Disclaimer — read this first. This plugin produces opinionated investment analysis for educational and informational purposes only. It is not personalized financial advice. Nothing produced by this plugin should be taken as a recommendation to buy, sell, or hold any security. Data is delayed ~15 minutes. Consult a qualified financial advisor before making investment decisions. Past performance does not predict future returns. Options involve substantial risk and can expire worthless. Leveraged ETFs are not long-term investments. You can lose money. Use at your own risk.

A Claude Code plugin for opinionated, aggressive-growth equity research — stocks, options, ETFs, sectors, and full portfolio reviews. Uses the Yahoo Finance MCP for quantitative data and web search for qualitative context.

What it does

  • Deep stock research — a 7-phase workup (business, sector, fundamentals, technicals, options, sentiment, verdict) culminating in a multi-horizon rating with explicit, quantified risk callouts for short / mid / long-term horizons.
  • Portfolio analysis — intake from Excel, CSV, screenshots, broker PDFs, or typed text. Normalized, confirmed back, then run through concentration, correlation, beta, drawdown, and factor analysis.
  • ETF analysis — methodology, holdings, factor exposure, correlation with your existing book, role fit.
  • Sector briefs — composition, aggregates, regime fit, best-in-class names, entry vehicles.
  • Options flow — IV rank, skew, put/call ratios, max pain, magnet strikes, implied moves, unusual activity.
  • Rebalance plans — concrete action matrices with triggers, deadlines, and risk callouts per line.
  • Idea generation — screens live market data for new stocks, ETFs, or options overlays to fill a mandate or a gap in your book — candidates are sourced from a screen and run through a real workup, never pulled from memory.

The framing

This plugin defaults to an aggressive growth, higher-than-average risk tolerance lens. It's willing to take a stance. It considers higher-beta names, thematic plays, leveraged ETFs (with warnings), and options overlays.

But every aggressive recommendation carries a mandatory, quantified risk callout — a WHAT KILLS THIS TRADE block specifying exactly what would invalidate the thesis and how much it would cost. Aggressive != reckless.

Install on Claude Code

This repo includes Claude Code plugin marketplace metadata. Install it with two commands from inside Claude Code:

/plugin marketplace add gmanish10/equity-research-skill
/plugin install equity-research@comprehensive-equity-research

After install, restart Claude Code to pick up the new slash commands.

Prerequisites

  • Claude Code with plugins enabled

  • Yahoo Finance MCP server configured

  • Python 3.10+ on the machine running Claude Code

  • Python packages: pandas numpy openpyxl rapidfuzz yfinance python-docx

    pip install pandas numpy openpyxl rapidfuzz yfinance python-docx
    

Install as a standalone skill

If you're not using a plugin-capable client, you can install just the skill (without the slash commands) by packaging or copying plugins/equity-research/skills/equity-research/ into your client's skill location. In clients that support .skill files, use the release asset if one is published.

Usage

Slash commands exposed by the plugin:

CommandWhat it does
/research-stock TICKERFull 7-phase deep dive
/analyze-portfolio [file]End-to-end portfolio review with intake
/analyze-etf TICKERETF one-pager
/sector-brief SECTORSector analysis + best-in-class picks
/options-flow TICKEROptions setup read
/rebalance-planAction matrix after a portfolio review
/find-ideas MANDATEScreen live data for new buy ideas (stocks, ETFs, overlays)

Or just ask naturally — the skill's description triggers automatically on investment-related queries:

  • "Do a deep research workup on NVDA. I want short/mid/long-term views."
  • "Here's my portfolio (attaches file). Review it with an aggressive growth lens."
  • "What's the options setup on TSLA for next month?"
  • "Give me a sector brief on semiconductors."
  • "Find me some new AI-exposure ideas that aren't already in my portfolio."
  • "Based on the review, what should I actually do this week?"

Directory structure

equity-research-skill/
|-- .claude-plugin/
|   \-- marketplace.json               # Claude Code marketplace manifest
|-- plugins/
|   \-- equity-research/               # Plugin root
|       |-- .claude-plugin/
|       |   \-- plugin.json            # Claude plugin manifest
|       |-- commands/                  # Slash commands
|       |   |-- research-stock.md
|       |   |-- analyze-portfolio.md
|       |   |-- analyze-etf.md
|       |   |-- sector-brief.md
|       |   |-- options-flow.md
|       |   |-- rebalance-plan.md
|       |   \-- find-ideas.md
|       \-- skills/
|           \-- equity-research/       # The skill
|               |-- SKILL.md           # Skill entry point
|               |-- references/        # Deep-dive methodology docs
|               |-- scripts/           # Python - technicals, ratios, portfolio intake, etc.
|               \-- evals/             # Eval test cases
|-- README.md
|-- LICENSE
|-- DISCLAIMER.md
\-- .gitignore

Credits

Built on top of Anthropic's bundled stock-analyst skill. The Yahoo Finance MCP orchestration patterns are adapted from that foundation. Specialization (multi-horizon ratings, aggressive-growth framing, ETF workflow, multi-format portfolio intake, risk-callout mandate) is original to this plugin.

Contributing

Open an issue or PR. Useful directions:

  • Additional broker-statement parsers in plugins/equity-research/skills/equity-research/scripts/parse_portfolio.py
  • More sophisticated DCF / fair value modeling
  • Backtesting harness for rating accuracy
  • Earnings-call transcript summarization
  • Alternative data integrations (e.g., short-seller report tracking)

License

MIT. See LICENSE.

One more time - this is not financial advice

The plugin is opinionated by design. That does not mean its opinions are correct. Do your own research. Consult a professional. Size positions appropriately. Know what you can afford to lose. Markets are adversarial and humbling.

Ver o original no GitHub

Bom para

Relacionados