AI-Trader

LLM-driven autonomous stock trading system with multi-agent architecture

Academic repo from HKU claiming profitable automated trading via LLM agents. Only two weak signals detected: rank 93 on GitHub (very low) and rank 12 on Bluesky (niche, tiny pool of 17 items). No coverage on HN, X, Reddit, Dev.to, or Product Hunt despite 2+ years of development. Divergence verdict is misleading—this is effectively single-source with Bluesky being statistically insignificant.

Why now: No genuine inflection—low-end GitHub placement coincided with tiny Bluesky pool inclusion creating artificial divergence signal in our composite.

• GitHub rank 93/100 in our pool—barely makes cutoff, no star velocity indicated
• Bluesky rank 12/17: pool has only 17 items, making this near-bottom performance disguised as top placement
• Zero signal on X, HN, Reddit, Product Hunt, Dev.to despite claims of profitable trading and 2+ year project lifespan
• Score of 1234599392 on GitHub is unnormalized/raw metric, not engagement—likely dependency on absolute star count without velocity

Considerations: Financial trading claims without verifiable live trading records or third-party audit are structurally suspect. The repo conflates backtested 'profitable' with real alpha; LLM agents for trading have failed replication across literature. Bluesky discourse is likely original author or team boosting, not organic. Missing from all high-credibility developer platforms suggests no practitioner adoption.

EMERGING SIGNAL · Ignore: Demand verified live trading Sharpe ratios or independent replication before allocating any attention; current signal profile matches academic promotion with zero real traction.

Sources: GitHub: HKUDS/AI-Trader · Bluesky search: HKUDS/AI-Trader