ginlix-ai/LangAlpha

Health v2 · maintenance only

Adoption not part of the score

1696 stars · 284 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

LangAlpha is an open-source agentic AI harness for financial market research and investment decision support, inspired by code-agent harnesses like Claude Code. It provides persistent workspaces, parallel research subagents, MCP tool integration, and access to market data such as prices, SEC filings, options chains, macro series, and news.

Use cases

• generate long/short pair-trade ideas from market screening
• build and iteratively refine an investment thesis with an AI agent
• analyze SEC filings and earnings call transcripts with citations
• run deep research on a sector rotation or macro theme
• track and update positions as new market data arrives
• automate scheduled market research workflows
• chat with an agent about stock fundamentals and options chains

When to choose

• you want an open-source, self-hostable AI analyst for equity research
• you need persistent, compounding research workspaces rather than one-shot Q&A
• you want agents grounded in primary sources like SEC EDGAR filings and earnings transcripts
• you want to extend an agent with custom MCP tools and skills for finance

When to avoid

• you need guaranteed-accurate financial advice or regulatory compliance
• you want a simple charting or portfolio tracker without LLM agents
• you need real-time HFT execution rather than research support
• you cannot work with an early-stage project requiring Python 3.13+

Facets

application · maturity active

agent-framework mcp rag web-scraping chat-interface data-visualization analytics trading llm-inference artificial-intelligence fintech large-language-models data-science python self-hosted cross-platform financial-research investment-agent vibe-investing langchain langgraph subagents market-analysis sec-filings equity-research trading-ideas ai-agents finance web-server

Member repositories

For agents

markdown · JSON · MCP: product_card(name="ginlix-ai/LangAlpha")