Practical overview
When it fits
Use it to build and revisit an ongoing personal equity-research process: start with market or sector signals, investigate companies, preserve supporting materials, and optionally use an agent for multi-step work or backtesting.
Who it is for
- Individual investors who want to organize equity research and market observations in one workspace.
- Researchers following A-shares, U.S. stocks, or Hong Kong stocks, with the documented full company-research workflow focused on A-shares.
- Users who already have access to a supported AI subscription or model API and want to apply it to research tasks.
Documented uses
- Review market movements and collect news, announcements, and event leads for follow-up.
- Track sector signals and companies, then develop a staged company analysis covering business, financials, earnings expectations, valuation, and risks.
- Import and search research documents, ask questions about their contents, and return to source text.
- Record holdings and watchlists, compare arguments in a research debate, and test defined rules against historical data.
Requirements & limits
The project is open source under the MIT license. AI access requires the user’s own supported subscription or model API; the project says subscriptions are not included.
- The agent is off by default; ordinary chat and agent-driven research are distinct modes.
- The full six-stage company-research workflow is documented for A-shares; other markets are supported for some records and features, including watchlists and holdings.
- Imported research files are saved and parsed locally, but matching text sent in a query goes to the selected model, so local storage does not mean fully offline processing.
- Portfolio or watchlist details extracted from an image or table are drafts that need manual review before saving.
- Backtests check historical rules; historical performance is not a prediction of future results or a real-world trading recommendation.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Connect an AI provider in Vibe Research
Configure a subscription or compatible model API for use across the research workspace.
Before you start
- A Claude Code, Codex, or WorkBuddy subscription, or a compatible Chat Completions model API.
Steps
Open the “Connect AI” section.
Choose a subscription or model API.
Test the connection; after a successful test, the connection is shared across the workspace.
Expected outcome
The selected AI connection is available across the workspace; the research agent is off by default.
Guide limitations
- The README does not establish that all providers offer identical behavior.
- The research agent must be enabled separately if you want to use it.
Run a conversational historical backtest
Check how an investment idea or rule performed against historical data.
Before you start
- A configured AI connection.
Steps
Open Research Tools and select Backtest.
Describe the idea or rule to check in the conversation.
Review the historical results against an index and the result charts.
Expected outcome
A historical backtest with index comparison and charts for cumulative return, drawdown, monthly results, and annual results.
Guide limitations
- The source does not specify supported rule syntax, historical-data coverage, or performance guarantees.
- Backtesting is for research; the portfolio tools do not place trades.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 9 October 2026
- Original website title
- simonlin1212/Vibe-Research · Official homepage title · retained in its source language