convez.ai
AgenteFinanceiroIntermediário
marras0914

Agent Toolbelt

marras0914/agent-toolbelt

Análise de ações sintetizada por IA + 20 ferramentas utilitárias para agentes

Ver no GitHub
InicialPluginScore · de 0 a 100

Conjunto de ferramentas especializadas para agentes de IA, com 7 ferramentas de análise de ações (dados financeiros ao vivo + síntese com IA) e 20 utilitários para infraestrutura de agentes. API em produção com modelo de pagamento por uso (1.000 chamadas/mês gratuitas).

Instalação

$npx convez-ai marras0914/agent-toolbelt

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
1
Forks
0
Acompanhando
0
Saúde
14%
Último commit
há 1 semana
Versão
v1.0.2
Issues abertas
0
Contribuidores
2

Casos de uso

  • Pesquisa de ações e teses de investimento automatizadas
  • Análise de earnings e sinais de insiders
  • Comparação de valuation entre múltiplas empresas
  • Extração de metadados, conversão de formatos e normalização de dados
  • Automação de tarefas comuns em workflows de agentes

Veredito

Toolbelt prático e bem desenhado para agentes que trabalham com finanças e processamento de dados. As ferramentas de stock research entregam análise sintetizada (não raw data), economizando tokens; os utilitários cobrem casos reais. Pricing justo e API clara, embora dependa de fontes terceiras (Polygon, Finnhub, FMP).

README

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

Agent Toolbelt

Stock research tools for AI agents. Live financial data + Claude-synthesized analysis, served as 7 focused tools — not raw OHLCV. Plus 20 utility tools for the rest of an agent's work.

Production API: https://www.agenttoolbelt.live


Quickstart

# Get a free API key (1,000 calls/month, no credit card)
curl -X POST 'https://www.agenttoolbelt.live/api/clients/register' \
  -H "Content-Type: application/json" \
  -d '{"email": "you@example.com"}'

# Generate a Motley Fool-style investment thesis for any ticker
curl -X POST https://www.agenttoolbelt.live/api/tools/stock-thesis \
  -H "Authorization: Bearer atb_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"ticker": "NVDA", "timeHorizon": "3-5 years"}'

Returns: bullish/neutral/bearish verdict, thesis paragraphs, key strengths, key risks, valuation read, insider read, analyst consensus read, and what to watch in the next earnings report.


Stock research tools (7)

LLM-synthesized analysis on top of live financial data from Polygon.io, Finnhub, and Financial Modeling Prep.

ToolWhat it doesPrice
stock-thesisFull Motley Fool-style investment thesis: verdict + thesis paragraphs + strengths + risks + valuation read$0.05/call
earnings-analysisEPS beat/miss history, revenue trend, long-term earnings consistency read, upcoming earnings date$0.05/call
insider-signalForm 4 interpretation — distinguishes meaningful open-market purchases from routine sales/awards. Signal strength + confidence$0.05/call
valuation-snapshotP/E, P/S, EV/EBITDA, FCF yield, ROE, margins → cheap/fair/expensive verdict + specific buy zone$0.05/call
bear-vs-bullSteelmanned 3-bull / 3-bear case with specific data, net verdict, key debate question$0.05/call
compare-stocksHead-to-head comparison of 2-3 tickers. Winner + per-ticker strengths/concerns + ifYouValue map (growth / value / quality)$0.05/call
moat-analysisBuffett-style competitive moat assessment (brand / switching costs / network / scale / IP / cost). Wide/narrow/none + durability$0.05/call

Every stock tool returns a dataSources block with fetchedAt + per-source success flags so you know exactly what data backed the analysis.


Utility tools (20)

Common agent infrastructure. Rule-based tools billed at $0.0001–$0.001/call; LLM-powered tools at $0.005–$0.10/call.

ToolWhat it doesPrice
text-extractorExtract emails, URLs, phones, dates, currencies, addresses, names from text$0.0005/call
token-counterCount tokens across 15 LLM models with cost estimates$0.0001/call
schema-generatorJSON Schema / TypeScript / Zod validator from plain English$0.001/call
csv-to-jsonCSV to typed JSON with auto delimiter and type casting$0.0005/call
markdown-converterHTML ↔ Markdown conversion$0.0005/call
url-metadataTitle, OG tags, favicon, author from any URL$0.001/call
web-summarizerFetch + summarize a URL with key points$0.02/call
regex-builderNatural language → regex with JS/Python/TS snippets$0.0005/call
cron-builderSchedule description → cron expression with next-run preview$0.0005/call
address-normalizerUS address → USPS format with component parsing$0.0005/call
color-paletteColor palettes with WCAG scores and CSS vars$0.0005/call
brand-kitFull brand kit — colors, typography, CSS/Tailwind tokens$0.001/call
image-metadata-stripperStrip EXIF/GPS/IPTC/XMP metadata for privacy$0.001/call
meeting-action-itemsAction items, decisions, summary from meeting notes$0.05/call
prompt-optimizerScore and rewrite LLM prompts$0.05/call
document-comparatorSemantic diff between two document versions$0.05/call
contract-clause-extractorKey clauses + risk flags from contracts$0.10/call
api-response-mockerRealistic mock data from a JSON Schema$0.0005/call
dependency-auditorCVE scan for npm/PyPI packages via OSV database$0.005/call
context-window-packerPack content into a token budget for LLM context$0.001/call

Do you know what your agent is actually calling?

You wired up the tools. Once the agent is running, though, you have no visibility into which ones it calls or what it passes to them. I also built Cordon for this. It sits in front of any MCP server and logs every call. You can write rules to block the ones you don't want touching production.

getcordon.com — npx @getcordon/cli init


npm SDK + LangChain

npm install agent-toolbelt

Typed client

import { AgentToolbelt } from "agent-toolbelt";

const client = new AgentToolbelt({ apiKey: process.env.AGENT_TOOLBELT_KEY! });

// Stock research
const thesis = await client.stockThesis({ ticker: "NVDA", timeHorizon: "3-5 years" });
const moat = await client.moatAnalysis({ ticker: "AAPL" });
const compare = await client.compareStocks({ tickers: ["NVDA", "AMD"] });

// Utility
const tokens = await client.tokenCounter({ text: myDocument });
const contacts = await client.textExtractor({
  text: emailBody,
  extractors: ["emails", "phone_numbers", "addresses"],
});

LangChain integration

import { AgentToolbelt } from "agent-toolbelt";
import { createLangChainTools } from "agent-toolbelt/langchain";
import { createReactAgent } from "@langchain/langgraph/prebuilt";
import { ChatOpenAI } from "@langchain/openai";

const client = new AgentToolbelt({ apiKey: process.env.AGENT_TOOLBELT_KEY! });
const tools = createLangChainTools(client); // 27 ready-to-use DynamicStructuredTools

const agent = createReactAgent({
  llm: new ChatOpenAI({ model: "gpt-4o" }),
  tools,
});

Claude MCP

Use the stock research tools (and the rest of the toolbelt) directly inside Claude Desktop or Claude Code via the agent-toolbelt-mcp package.

Claude Desktop — add to claude_desktop_config.json:

{
  "mcpServers": {
    "agent-toolbelt": {
      "command": "npx",
      "args": ["-y", "agent-toolbelt-mcp"],
      "env": {
        "AGENT_TOOLBELT_KEY": "atb_your_key_here"
      }
    }
  }
}

Claude Code — one command:

claude mcp add agent-toolbelt -e AGENT_TOOLBELT_KEY=atb_your_key_here -- npx -y agent-toolbelt-mcp

Once installed, ask Claude things like "Give me a full analysis of NVDA — thesis, earnings quality, insider activity, and whether it's cheap right now" and it'll call the tools in parallel.


Discover tools programmatically

Agents can auto-discover all 27 tools at runtime:

curl https://www.agenttoolbelt.live/api/tools/catalog

Pricing

TierPriceMonthly callsRate limit
Free$0/mo1,00010/min
PAYGprepaid creditsunlimited60/min
Starter$29/mo50,00060/min
Pro$99/mo500,000300/min
EnterpriseCustom5,000,0001,000/min

Integrations

  • npm SDK — npm install agent-toolbelt — typed client + LangChain tools
  • MCP — npx -y agent-toolbelt-mcp — works with Claude Desktop and Claude Code
  • OpenAI GPT Actions — OpenAPI spec at /openapi/openapi-gpt-actions.json
  • RapidAPI — listed on the RapidAPI marketplace
  • Smithery, Glama, PulseMCP, MCP registry — discoverable in MCP directories

Going to production

When the agent moves out of dev, a new set of questions shows up. What did it call last night? What arguments did it pass? Who approved the destructive one?

Cordon is an MCP gateway that sits in front of servers like this one. Point your client at Cordon instead of directly at Agent Toolbelt; Cordon forwards every call through and adds:

  • A real-time audit log of every tool invocation (name, arguments, response, latency)
  • Per-API-key policy: which tools each caller can use, under what conditions
  • Slack-based human approvals for tool calls you've flagged as high-risk

From the agent's perspective nothing changes — same tools, same schemas. Free tier covers 1,000 events/month.


License

MIT

Ver o original no GitHub

Bom para

Relacionados