FAQ
Is this a stock recommendation service?
No. This is a valuation discipline tool that helps you assess whether a stock is expensive and what its risk-reward profile looks like, using an institutional investment-committee methodology grounded in source-bound evidence. It does not predict price movements or issue buy/sell instructions. It is an educational research tool, not investment advice.
Are the reports reliable? Could the engine fabricate data?
The engine operates under hard constraints: it must not fabricate prices, EPS, news, SEC/IR facts, consensus estimates, historical or terminal multiples, or DCF inputs. When data cannot be retrieved, it is flagged as missing and confidence is lowered. Provisional and Formal outputs are strictly distinguished.
Who owns my reports and cards?
You do. Reports can be saved as a standard Locked Card, stored under your account or written to your own GitHub repository, and taken with you at any time.
What are Data Refresh and Revision?
Data Refresh recomputes the current conclusion against new price and fundamentals without altering long-term assumptions. Revision is triggered when evidence indicates that long-term assumptions have changed: the valuation is redone and a new card version is generated. Both build on your existing Locked Card.
Will you have access to all of my GitHub code?
We only write to the single repository you explicitly bind, and never read from or write to any other repository. Authorization is controlled by you on GitHub's side and can be revoked at any time.
Which AI models do you use?
The engine supports five interchangeable model providers (OpenAI / Claude / Gemini / DeepSeek / Qwen). The free Snapshot tier uses cost-efficient models, while the paid Institutional Locked Card uses the strongest configured model to ensure faithful rule execution.