LeoCode by LeonardXLeoCode is a native Visual Studio Code extension project that connects to the existing LeonardX account and Leo chat APIs. It does not bundle an AI provider key or run commands generated by the model. Current functionality
The model picker uses five LEO-branded configurations backed by distinct Google Gemini model IDs. The upstream model identifiers are cross-checked against the official Gemini models catalog; actual availability for a particular API key is determined by the server's provider response, and no credentials are included in the extension. The chat API currently returns a completed JSON response rather than a token stream, so this extension does not claim streaming. Founders Early Access accounts share a server-enforced limit of 40 Leo chat and sandbox requests per rolling 24-hour window. The entitlement and usage counter are checked by LeonardX; the extension does not store or enforce the quota locally. Active editor context is sent to LeonardX when you submit a chat or specialist-agent request. LeoCode excludes common credential files and redacts common secret assignments, but this is a best-effort safeguard, not a guarantee that every sensitive value is detected. Review the active editor and selection before sending. Generated code is shown as a diff and is not applied unless you confirm. Local development
The default server origin is Install from the MarketplaceLeoCode is published by the verified After installation, select the LeoCode icon in the Activity Bar, open Leo Chat, and choose Connect to sign in to your LeonardX account. |