Agents & Coding

Vibe-Trading

Conversational quant research, with kill switches

Open ✓

What it is

From HKU's data-science lab: natural-language quantitative research that turns a hypothesis into a signal engine and backtest. Ships 50+ MCP tools, an 'Alpha Zoo' of 452 prebuilt factors, 18 free data providers, connectors to 10 brokers, and delivery to 16 chat channels — with mandate gates, kill switches, and audit ledgers built in.

Why it's interesting

Makes institutional-style factor research conversational, using strict point-in-time data to avoid lookahead bias. And because it wires agents to live broker accounts, its safety rails are as editorially interesting as its features.

Use cases

  • Natural-language backtesting and factor research
  • Automated research digests to Telegram or Slack
  • Paper-trading agent experiments

Who it's for

Quant-curious developers, retail algo traders, researchers

Setup

Advanced. Python 3.11+, Docker, an LLM API key; broker and data credentials for live use

Limitations & cautions

An LLM agent with write access to a real brokerage account deserves maximum caution: paper-trade first, keep mandates and kill switches on, and remember backtests are not future returns. Nothing here is financial advice, and the regulatory obligations are yours.

Editorial takeaway

The most instructive part isn't the alpha factory — it's watching a serious lab design guardrails for agents that touch money.

Related & alternatives