← Back to Catalog

RKiding/Awesome-finance-skills

↗ GitHub

A collection of Awesome Finance Agent Skills for free and easy to start | 一系列开源免费的金融分析Agent Skills

1,556

Stars

200

Forks

6

Watchers

0

Open Issues

Python·Apache License 2.0·Last commit Mar 29, 2026·by @RKiding·Published April 3, 2026·Analyzed 5d ago
A

Safety Rating A

No hardcoded secrets, malicious code patterns, suspicious dependencies, or prompt injection attempts were detected. The repository is a collection of SKILL.md-based agent skill definitions and Python tooling for financial analysis. The README is straightforward documentation with bilingual (English/Chinese) content. The project is Apache 2.0 licensed and links to legitimate related projects. No red flags were found.

AI-assisted review, not a professional security audit.

AI Analysis

Awesome Finance Skills is a plug-and-play collection of AI agent skill modules focused on financial analysis. Each skill is a structured SKILL.md-based module that equips LLM-powered agents with capabilities including real-time financial news aggregation from 10+ sources (Cailian, WSJ, Weibo, Polymarket), stock data retrieval for A-Share/HK/US markets, FinBERT/LLM-based sentiment analysis, Kronos time-series forecasting with news-aware adjustments, investment signal tracking, logic chain/transmission diagram visualization (Draw.io XML), professional financial report generation, and web/RAG search. Skills are compatible with multiple agent frameworks including Claude Code, OpenCode, Antigravity, and OpenClaw, and can be installed via npx or manually copied to framework skill directories.

Use Cases

  • Equipping AI coding agents (Claude Code, OpenCode, Codex) with financial analysis capabilities via plug-and-play skill modules
  • Real-time aggregation and analysis of financial news from Chinese and global sources
  • Retrieving stock market data (OHLCV, fundamentals) for A-Share, Hong Kong, and US equity markets
  • Running sentiment analysis on financial text using FinBERT or LLM-based scoring
  • Forecasting stock price movements using the Kronos time-series model with news-sentiment adjustments
  • Generating professional financial research reports through a plan-write-edit-chart pipeline
  • Visualizing market logic transmission chains as Draw.io diagrams
  • Tracking evolution of investment signals (strengthen/weaken/falsify)

Tags

#ai-agents#llm#plugin#rag#analytics#api-wrapper#workflow-automation#library#prompt-management#context-engineering#research#data#framework

Project Connections

Alternative to

TauricResearch/TradingAgents

Both projects provide LLM-powered financial analysis capabilities. TradingAgents uses a multi-agent LangGraph framework for trading research, while Awesome Finance Skills provides plug-and-play skill modules for agent frameworks like Claude Code. They solve similar finance analysis problems via different architectural approaches.

Complements

OpenBB-finance/OpenBB

OpenBB provides the underlying financial data infrastructure (equities, crypto, economics) via Python SDK, REST API, and MCP servers. Awesome Finance Skills could use OpenBB as a data source backend for its stock and market data skills.

Alternative to

garrytan/gstack

Both projects are collections of SKILL.md-based agent skills installable into Claude Code and similar agent runtimes. gstack focuses on software engineering team roles, while Awesome Finance Skills focuses on financial analysis domain skills. They share the same skill packaging and distribution pattern.

Complements

google-research/timesfm

TimesFM is a pretrained time-series forecasting foundation model. The alphaear-predictor skill in Awesome Finance Skills uses the Kronos forecasting model for stock prediction; TimesFM could serve as an alternative or complementary forecasting backend.

Complements

One-Man-Company/Skills-ContextManager

Skills-ContextManager is an MCP server and web UI for organizing and dynamically loading agent skills. Awesome Finance Skills provides a domain-specific skill collection that could be managed, organized, and served via Skills-ContextManager to any MCP-compatible agent.