Feature request: Add Xianyu (闲鱼) as a searchable channel for real-market demand & pain-point analysis

[anthonyhann]

Feature request: Add Xianyu (闲鱼) as a searchable channel to support real-market demand & pain-point analysis

Hi, thanks for the great tool! Agent Reach already covers Xiaohongshu, Twitter, Reddit, Bilibili, etc. I'd like to propose adding Xianyu (闲鱼, Alibaba's leading second-hand / C2C marketplace in China) as a new searchable channel.

Why Xianyu matters for market analysis

• Xianyu is one of the largest C2C second-hand platforms in China (hundreds of millions of users). It is where real, unfiltered demand and pain points surface: what people actually want to buy/sell, at what price, and what complaints they voice.
• For anyone doing market research, demand validation, or pain-point discovery (e.g. niche products, local services, cross-border goods), Xianyu listings + comments are a goldmine of authentic, transaction-level signal that search engines and social platforms don't capture well.
• It would give "demand and pain points" a real market-analysis data backing rather than relying only on social-media sentiment.

What I'd like

• Ability to search Xianyu listings (keywords, category, price range) and read listing details / seller descriptions / comments via Agent Reach.
• Preferably via the existing OpenCLI-style browser backend (like Xiaohongshu / Reddit) since Xianyu requires a logged-in Taobao/Alipay session and is anti-bot heavy.
• A documented adapter name (e.g. opencli xianyu search "<keyword>") consistent with the current routing.

Suggested approach / considerations

• Xianyu has strong anti-crawling and requires an authenticated session (Taobao/Alipay login). A headless / OpenCLI bridge reusing the user's browser login is the most practical path.
• Rate limiting and login-state handling should follow the same "no auto-login, credentials stay local" safety rules already in the SKILL docs.

Pain points to solve for users

• Manual Xianyu research is slow and hard to aggregate.
• No easy way to turn listing / comment data into structured market signals.

Would love to hear your thoughts on feasibility and whether a browser-backend adapter (similar to the existing xiaohongshu/reddit ones) is the right direction. Happy to help test if a prototype lands.

Thanks!