Show HN: LangAlpha – what if Claude Code was built for Wall Street?
Claude Code for Wall Street just went open source. Here's why financial agents need persistent memory, not one-shot PDFs.
Why it matters
LangAlpha solves a critical gap in AI-powered financial research: multi-session persistence and domain context. Most agent frameworks treat deliverables as endpoints; professional investing requires iterative analysis that compounds across weeks. This open-source tool addresses a real workflow problem that existing AI investing platforms miss.
The key facts
7 to knowOpen-sourced financial agent framework built on Claude
Auto-generates typed Python modules from MCP schemas to reduce token overhead (50k+ token savings per server)
Persistent workspace model maintains research context across sessions with agent memory files
~80 tools integrated with unified prompt cost regardless of tool density
Injected domain context (portfolio, watchlist, risk tolerance, financial data sources) into every LLM call
Addresses iterative research workflow (earnings updates, competitive analysis, layered research) vs. one-shot deliverable model
Published on Hacker News with 31 points, 12 comments
Go to the source
Hacker Newsgithub.com
Publisher excerpt: Some technical context on what we ran into building this. MCP tools don't really work for financial data at scale. One tool call for five years of daily prices dumps tens of thousands of tokens into the context window. And data vendors pack dozens of tools into a single MCP server, schemas alone…