Integrate with Gemini CLI

Use Gemini CLI as a coding assistant to build the Spendkit integration in your Agent project. The running Agent later authenticates to Spendkit and pays through its bound wallet.

Add project context

# GEMINI.md
For AI Agent payment work, read:
https://spendkit-alpha.vercel.app/docs/ai-coding-agents/
https://spendkit-alpha.vercel.app/docs/runtime/
https://spendkit-alpha.vercel.app/docs/examples/
https://spendkit-alpha.vercel.app/docs/api-quickstart/
https://spendkit-alpha.vercel.app/docs/mcp-tools/

Inspect MCP discovery, OAuth Client Credentials refresh, and Payment Wallet submission.
Never request the Payment Wallet private key in chat or put it in source control.
Do not create or expand token allowance or change Policy limits without the owner.
Implement and inspect the read-only connection and preview before enabling payment.

Authentication is part of the Runtime

Spendkit exchanges the Agent connection's Client ID and Client Secret for a short-lived Bearer token. Do not place one temporary access token into permanent MCP settings. Your Runtime must be able to obtain a new token when needed, use the live MCP tool schemas, and submit payment transactions from the verified wallet separately from the model.

See the generic MCP requirements → Review the wallet boundary →