Skills— plugin

<div align="center">

<img src="assets/llmquant-skills-logo.svg" alt="LLMQuant Skills" width="128" />

<h1>LLMQuant Skills</h1>

<p><strong>Reusable finance Agent Skills — grounded in <a href="https://github.com/LLMQuant/data-mcp">LLMQuant Data</a></strong></p>

<p><strong>English</strong> · <a href="README.zh-CN.md">简体中文</a></p>

<p> <a href="https://github.com/LLMQuant/skills/stargazers"><img src="https://img.shields.io/github/stars/LLMQuant/skills?style=flat" alt="GitHub stars" /></a> <a href="LICENSE"><img src="https://img.shields.io/badge/License-MIT-lightgrey.svg" alt="License: MIT" /></a> <a href="https://github.com/LLMQuant/data-mcp"><img src="https://img.shields.io/badge/native%20data-LLMQuant%20Data%20MCP-0B5FFF" alt="Native LLMQuant Data MCP" /></a> <a href="https://llmquantdata.com/agent"><img src="https://img.shields.io/badge/agent%20playground-open-8A5A44" alt="Open LLMQuant Agent Playground" /></a> <a href="https://github.com/LLMQuant/skills/commits"><img src="https://img.shields.io/github/last-commit/LLMQuant/skills" alt="Last commit" /></a> </p>

</div>

npx skills add LLMQuant/skills      # Claude Code · Codex · Cursor · Antigravity · OpenClaw · Hermes · …

[!TIP] 18 category skills — equities, options, macro, crypto, credit, portfolio, risk, and more — that route an agent to the right finance workflow and keep every claim grounded in LLMQuant Data.

Contents

• Contents
• Overview
• Category Skills
• Install
• Recommended: npx skills add
• Native plugin install
• LLMQuant Data
• Contributing
• License
• Star History

Overview

<div align="center"> <img src="assets/llmquant-skills-header.png" alt="LLMQuant Skills — finance Agent Skills grounded in LLMQuant Data" width="820" /> </div>

This repository is a skill catalog. The install/import unit is a category folder under skills/llmquant-* — each one a router SKILL.md that indexes workflows/*.md, tells the agent which workflow to load, and enforces the LLMQuant Data evidence contract. It is not an isolated SKILL.md and not a single workflow file.

<details> <summary><strong>Repository layout</strong></summary>

skills/
├── llmquant-data/
│   ├── SKILL.md
│   ├── workflows/
│   ├── scripts/
│   └── assets/
├── llmquant-etfs/
│   ├── SKILL.md
│   └── workflows/
└── ...

.claude-plugin/     # Claude Code plugin + marketplace manifests
.codex-plugin/      # Codex plugin manifest
.cursor-plugin/     # Cursor plugin manifest
assets/
README.md
README.zh-CN.md

</details>

Category Skills

Install

Recommended: npx skills add

One command, and it works across Claude Code, Codex, Cursor, Antigravity, Gemini, and other agents.

[!TIP] Already inside an agent? Just run npx skills add LLMQuant/skills — it auto-detects the host and installs to the right place.

# Pick skills interactively for the current agent
npx skills add LLMQuant/skills

# Install all category skills globally
npx skills add LLMQuant/skills -g --all

# Install specific skills
npx skills add LLMQuant/skills -g --skill llmquant-options llmquant-equities

# Target a specific agent
npx skills add LLMQuant/skills -a codex

Manage installed skills with npx skills list, npx skills update, and npx skills remove. Then ask the agent What skills are available?, or invoke one directly, e.g. /llmquant-options.

<div align="center"> <img src="assets/skills-add-screenshot.png" alt="Selecting LLMQuant skills with npx skills add" width="720" /> </div>

Native plugin install

The whole repository is also packaged as a single plugin that bundles every category skill, for agents with a native plugin system.

Claude Code — add this repo as a plugin marketplace, then install the bundle:

/plugin marketplace add LLMQuant/skills
/plugin install llmquant-skills@llmquant

Codex — the repo ships .codex-plugin/plugin.json, so it is plugin-ready. To install today, use the CLI above (npx skills add LLMQuant/skills -a codex), or install a single category skill via the skill installer:

$skill-installer install https://github.com/LLMQuant/skills/tree/main/skills/llmquant-options

Cursor · Antigravity · other agents — use the recommended CLI with the matching agent, e.g. npx skills add LLMQuant/skills -a cursor or -a antigravity.

LLMQuant Data

These skills are the workflow layer for [LLMQuant Data](https://github.com/LLMQuant/data-mcp) — one MCP server that feeds prices, filings, 13F, macro, ETF holdings, crypto, and more to your agents. Configure once, and every agent in your stack gets the data.

<div align="center"> <img src="assets/ecosystem.png" alt="LLMQuant Data, everywhere your agent runs — Claude Code, Cursor, Codex, Gemini, and more" width="800" /> </div>

Set it up the AI-native way — paste this to your agent:

Install the LLMQuant data-mcp server in this environment by following https://github.com/LLMQuant/data-mcp

Or add it by hand:

claude mcp add llmquant-data -e LLMQUANT_API_KEY=your_api_key -- npx -y @llmquant/data-mcp

[!NOTE] Grab an API key and the full multi-agent setup at [docs.llmquantdata.com](https://docs.llmquantdata.com) and [`LLMQuant/data-mcp`](https://github.com/LLMQuant/data-mcp).

Skills describe the data they need in natural language and let the agent route to whatever LLMQuant Data exposes, so coverage can grow without editing any skill. Without llmquant-data connected, skills still run as reusable workflows — the agent asks for user-provided data and clearly labels anything missing. The full evidence contract is in CONTRIBUTING.md.

<div align="center"> <a href="https://llmquantdata.com/agent"><img src="assets/playground.png" alt="LLMQuant Agent Playground — Data MCP and Skills enabled" width="800" /></a> <br/> <strong><a href="https://llmquantdata.com/agent">Open the LLMQuant Agent Playground →</a></strong> </div>

<details> <summary><strong>▶ Watch the playground demo</strong></summary> <br/> <div align="center"> <img src="assets/llmquant-data-agentplayground-demo.gif" alt="LLMQuant Agent Playground demo — Data MCP and Skills in action" width="800" /> </div> </details>

Contributing

Add or update workflows inside the relevant category (skills/llmquant-<category>/ → SKILL.md + workflows/ + scripts/ + assets/):

• Category folders must be named llmquant-*, and SKILL.md is the router that indexes every workflow.
• Workflow files hold the repeatable procedure, output format, data contract, and guardrails.
• Describe data needs as natural-language capabilities, not exact MCP tool names.
• Update this README when a category or major workflow is added, removed, or renamed.

See CONTRIBUTING.md for the full contract and quality bar.

License

MIT — see LICENSE. Listing a project here does not change the license of any linked code or artifact.

<div align="center"> <br/> <img src="assets/llmquant-logo.svg" alt="LLMQuant" width="64" /> <br/><br/> <strong><a href="https://llmquant.com">LLMQuant</a></strong> <br/> _{Open-source community for AI, LLMs, and quantitative finance.} <br/><br/> <a href="https://llmquant.com">Website</a> · <a href="https://github.com/LLMQuant">GitHub</a> · <a href="https://linkedin.com/company/llmquant">LinkedIn</a> </div>

Star History