Claude Code skills for investment analysis — options flow scanner + price alerts (open source)

[tellmefrankie]

Hi AutoGen community,

Multi-agent systems builders here — sharing an open source project that demonstrates agentic workflows for investment analysis.

AI Investment Skills — Claude Code skills for market monitoring:

• Options flow scanner: Detects unusual sector ETF P/C ratios, filters out noise (lottery tickets, micro-priced hedges)
• Price monitor: Scheduled agent that checks stop-loss/take-profit levels and sends Telegram alerts
• Daily briefing: Morning AI market summary

Multi-agent angle:
The options scanner pipeline has multiple stages: (1) data fetcher agent, (2) volume quality filter, (3) ratio analyzer, (4) anomaly detector, (5) alert dispatcher. Each stage could be a separate AutoGen agent. The current implementation is single-agent but the decomposition is natural.

Open source:
→ https://github.com/tellmefrankie/ai-investment-skills

The signal that validated the approach: XLI P/C hit 5.32 (normal: 0.5–1.2). After filtering: 98% was $0.01 lottery calls on CEG. Filtered signal: "extremely bearish." AutoGen users interested in finance domain applications?