# HKUDS/Vibe-Trading

"Vibe-Trading: Your Personal Trading Agent"

Repository: https://github.com/HKUDS/Vibe-Trading
Canonical: https://ross.abutalabs.com/products/vibe-trading
Homepage: https://vibetrading.wiki/
Language: Python
License: MIT
License Family: permissive
Topics: backtesting, multi-agent, quantitative-finance, trading, ai-agent, algorithmic-trading, fintech, llm, mcp, python
Last push: 2026-08-26T13:27:10+00:00

## Health v2 (maintenance only)
Score: 81/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 98, longevity 11
- inputs: {"age_days": 154, "days_push": 7, "days_rel": 14, "gap_med": 11.0, "n_releases_24m": 11}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 31791, forks 5175 (observed 2026-08-28T04:11:57.087408+00:00)

## What it is
Vibe-Trading is an open-source AI finance research agent that turns natural-language prompts into runnable market research, backtests, and multi-agent swarm analyses. It provides a CLI and MCP/API server with seven backtesting engines, persistent agent memory, and reproducible research artifacts across equities, crypto, futures, forex, and options.

## Use cases
- backtest a moving-average strategy on BTC-USDT from a natural language prompt
- run a multi-agent investment committee analysis on a stock
- parse my broker trade journal and run counterfactual backtests
- build an AI agent with trading and market data capabilities via MCP
- do quantitative factor analysis on crypto and equities
- generate reproducible finance research reports with evidence and caveats

## When to choose
- you want LLM-driven finance research and backtesting rather than live brokerage execution
- you need multi-agent swarm workflows like quant desk or risk review teams
- you want a CLI or MCP server that plugs trading research skills into your own agent

## When to avoid
- you need actual order execution or brokerage integration - it explicitly does not execute trades
- you require guaranteed-accurate financial advice; outputs are research artifacts with caveats
- you need a non-Python stack or a managed SaaS product

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, llm-inference, mcp, benchmarking, data-science, cli
- domain: fintech, artificial-intelligence, large-language-models, data-science, analytics
- platform: python, cli, cross-platform
- tags: trading, backtesting, quantitative-finance, multi-agent, finance-research, algorithmic-trading, shadow-account, swarm-teams, ai-agents

## Member repositories
- HKUDS/Vibe-Trading (main) score 81

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:57.087408+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:52:14.385483+00:00, confidence not recorded.
  - readme: https://github.com/HKUDS/Vibe-Trading (fetched 2026-08-28T04:11:57.087408+00:00, sha 33bc417a71cf)
  - homepage: https://vibetrading.wiki/ (fetched 2026-08-29T07:49:12.552653+00:00, sha 10effefd0052)
  - site_page: https://vibetrading.wiki/docs (fetched 2026-08-29T07:49:12.562683+00:00, sha 060a2f373e9e)
- Data as of 2026-08-30T08:39:29.467469+00:00.