Your key stays local. The code stays open.
TradeApe is built as a BYOK, open-source market analysis workspace: no bundled model credits, no hosted agent backend, and no black box between you and the code.
API keys are local to your browser.
TradeApe does not bundle hosted model access, agent credits, or a backend agent service. You provide your own Anthropic API key when you want the agent chat layer to run.
In the current app, that key is stored in your browser's localStorage. It is not sent to a TradeApe server. It is used from your session to call the model provider you configured.
The project is auditable.
TradeApe is open source so users and contributors can inspect how keys, prompts, chart tools, market data requests, and rendered overlays work. The goal is a financial analysis toolchain that can be read, forked, challenged, and improved in public.
If you find a privacy, security, provenance, or analysis-calibration issue, contributions and review are welcome through the GitHub project.
Market-data provenance is part of the product.
TradeApe shows provider, feed, symbol mapping, schema, capability, and candle-state details when available. Unknown exchange attribution is treated as a limitation, not filled in with guesses.
The default demo path uses QuestDB demo data. Volume and reference VWAP should be treated as feed-local context unless a named venue and explicit session anchor are configured.
This is not financial advice.
TradeApe is charting and analysis software. It can summarize deterministic tool output, draw candidate levels, and explain market structure evidence, but it does not provide trading signals, recommendations, execution guidance, or investment advice.
Any market read should be treated as provisional software output that requires independent validation. You are responsible for your own research, risk decisions, and use of any market data or model provider.