Wind Alice Investment Idea Generation

Problem Solved

When investment screening still relies on manual announcement review, factor spreadsheets, and company-by-company valuation work, Wind Alice Investment Idea Generation hands the market-wide candidate discovery step to the Wind Alice Agent. It is aimed at active investment idea discovery, helping users identify value, growth, quality, short, and event-driven candidates across A-shares, HK-listed names, US stocks, or global markets, rather than issuing trade instructions.

How It Works

• Prompt-driven screening: the agent assembles market, sector, theme, and factor conditions into a natural-language request, then invokes the service to scan the market.
• Multiple research modes: it supports quant factor screening, thematic scanning, and combined approaches, such as low valuation with high dividends, revenue and earnings growth, ROE and cash-flow quality, or beneficiaries of AI, new energy, and consumption upgrade themes.
• Peer benchmarking: it can build a peer set and provide relative valuation references such as PE, PB, and EV/EBITDA to help test mispricing logic.
• Structured delivery: the output typically lists no more than 10 stocks, prioritizes 3–5 core ideas, and frames each idea with non-consensus insight, catalysts, and key risks; user-facing delivery should use agentResult.value from the CLI stdout instead of pasting silently downloaded report attachments.

Scope and Caveats

This skill is best used for research shortlisting and idea supplementation, not as a replacement for detailed diligence, compliance checks, or independent judgment. It requires WIND_API_KEY configuration, and in multi-agent environments session isolation should be considered. Before running parallel work, use check-conflict to decide whether to attach, create a new task, or cancel.