clawock

Based on the available evidence, clawock is best judged as a promising open-source experiment rather than a broadly adopted investment research product. What the evidence supports is fairly specific: the official GitHub repo describes a multi-agent LLM workflow that runs daily bull-vs-bear debates on a real HK and US stock portfolio, applies hard risk gates, and keeps a daily scorecard of its own calls. That shows a clear methodology and an auditable intent, but the evidence sample is only one GitHub source, so neither adoption nor effectiveness should be overstated.

In practice, this is not an auto-trading bot and not a traditional quant backtesting platform. A more accurate analogy is a reproducible research workflow that turns “morning meeting”-style investment debate, risk constraints, and post-hoc grading into an open-source pipeline. For users, the output looks more like daily analysis artifacts and performance tracking than a guaranteed alpha engine. The repo snippet supports the claim that it has a live dashboard and self-evaluation loop, but that does not by itself prove statistically reliable forecasting quality.