convez.ai
SkillFinanceiroAvançado
mikepesh

Forced Holding Screen

mikepesh/forced-holding-screen

Identifica exposição forçada a IPOs em seus fundos e sugere swaps equivalentes

Ver no GitHub
InicialPluginScore · de 0 a 100

Skill que identifica quais fundos de índice forçariam você a manter ações de empresas específicas (como SpaceX, Anthropic ou OpenAI após IPO) e sugere alternativas limpas equivalentes. Usa busca web em tempo real para verificar status de IPOs e composição de fundos, processando apenas tickers e descrições (sem dados sensíveis).

Instalação

$npx convez-ai mikepesh/forced-holding-screen

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
—
Issues abertas
0
Contribuidores
1

Casos de uso

  • Planejar portfólio antes de IPOs disruptivos
  • Encontrar alternativas de fundos sem exposição a empresas específicas
  • Analisar risco de alocação forçada em diferentes tipos de conta (IRA, 401k, taxável)

Veredito

Ferramenta nicho bem-executada para investidores sofisticados que querem evitar exposição involuntária a IPOs. Design inteligente de privacidade (entrada de-identificada) e uso de web search para dados sempre atualizados. Advertências claras sobre não ser aconselhamento financeiro.

README

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

Forced-Holding Screen

license: MIT

A Claude skill that finds out where your index funds would force you to hold a specific stock — and what currently-clean alternatives cover the same market exposure without it.

When a big private company IPOs (SpaceX, and eventually Anthropic / OpenAI), it doesn't just become buyable — it gets added to indexes, and then every holder of those index funds is forced to own it whether they want to or not. This skill maps that risk across your portfolio, explains the at-risk holdings, and suggests swaps that keep your exposure but skip the name.

It runs on de-identified input by design: you paste only ticker + description (and optionally a percentage), so no dollar amounts, share counts, or account numbers are ever shared.

[!WARNING] Not investment, tax, or legal advice. No warranty. This is informational tooling. Suggested "clean equivalent" funds are examples of a category, not recommendations. Fund holdings and index rules change constantly — verify against primary sources before acting.

What it does

  • Buckets every holding by forced-holding risk: holds now (funds with a stake today), imminent (total-market and Nasdaq-100 funds pick up new listings fast), buffered (the S&P 500's profitability screen keeps brand-new IPOs out near-term), structurally safe (value, international, bonds), and individual stocks (no forced risk).
  • For at-risk holdings, explains why, suggests a currently-clean equivalent, what you keep vs. trade off, and — if you note the account type — whether swapping is tax-free (IRA/401k) or a taxable event.
  • Uses web search at run time to check current IPO status and fund holdings, so it isn't relying on stale baked-in data.
  • Generalizes to any company you want to avoid — not just the three above.

See examples/sample-report.md for what the output looks like, and examples/sample-input.txt for the paste format.

Install

This is a Claude skill — a single SKILL.md. Drop the folder into your skills directory:

  • Claude Code / Cowork: place forced-holding-screen/ in your skills folder (e.g. ~/.claude/skills/ or your project's skills/ — see Anthropic's skills docs for your setup).
  • Then just ask Claude something like "check my portfolio for forced exposure to the SpaceX IPO" and it triggers.

No dependencies, no install scripts, nothing to run.

Use

  1. Export your positions (in Fidelity: Accounts → Positions → Download).
  2. Paste only Symbol + Description (optionally "% of Account" and account type). Leave out dollar amounts, quantities, and account numbers.
  3. Claude verifies current facts, classifies each holding, and returns a report with suggested clean swaps.
  4. To "monitor," re-run it after an IPO — the ad-hoc version of a tripwire.

Why a skill (and not a script)

Because the input is de-identified, there's nothing sensitive to protect, so it doesn't need to run locally. And because index membership and fund holdings change constantly, live web search beats any data file that would go stale. The result is one portable text file instead of an app to install and maintain.

Built with AI

This skill was designed and specified by Mike "Pesh" Poeschl and built using Claude (Anthropic) as the primary authoring tool. The requirements, logic, privacy-by-design constraints, and classification framework came from the author; Claude translated those specifications into the SKILL.md prompt and supporting files.

It is not hand-coded in the traditional sense — it is a human-directed, AI-assisted artifact. That distinction feels worth naming: the judgment calls and design decisions are human; the drafting is AI. If you use or adapt this, you're welcome to do the same.

Contributing

Issues and PRs welcome — especially clarifications to the classification logic or the clean-equivalent guidance in SKILL.md. Keep it informational (no buy/sell advice) and keep the de-identified-input principle intact.

License

MIT — see LICENSE.

Ver o original no GitHub

Bom para

Relacionados