Build and publish a BaZi chart reading agent

Tools

Codex

Overview

I built an AI fortune-telling agent and it's now LIVE on @OKX AI Marketplace, earning USDT on-chain.

Workflow

The author built a BaZi (Four Pillars) fortune-telling agent named Destiny Codex using AI coding agents. The system uses a deterministic calendar engine to compute charts, with AI interpreting the data. It offers four services (horoscope, life reading, compatibility, date picking) with on-chain USDT payments. The author handled the entire build solo, from spec to live listing, in under a week.

Reported result

The agent is live on the OKX AI Marketplace and has settled over 20 real transactions. It supports multiple languages and generates shareable chart images. The system runs autonomously 24/7 with on-chain settlement. No further performance metrics or user feedback details are provided in the source.

Human role

The original post does not say how a person stayed involved.

Limitations

The source does not specify human oversight protocols during operation or maintenance. Integration with the OKX Marketplace and X Layer blockchain requires specific technical permissions and wallet setups. The deterministic chart engine must be accurately implemented to avoid hallucinations. Pricing and service availability are subject to marketplace rules.

Try it

Act as a developer building a BaZi fortune-telling agent for the OKX AI Marketplace. Use AI coding tools to create a deterministic chart engine that calculates Four Pillars based on solar terms and stem-branch calendars. Implement four services: Daily Horoscope (0.1 USDT), Full Life Reading (1 USDT), Couple Compatibility (0.5 USDT), and Auspicious Date Picker (0.3 USDT). Integrate on-chain USDT payments via X Layer. Ensure the AI interprets only computed data, not hallucinated charts. Support Chinese, English, and Vietnamese. Before publishing, run a sample calculation for a known birth date and verify the chart accuracy. Confirm all payment flows work in testnet. Only proceed to mainnet after human review of the code and test results. Output the complete code structure and deployment steps.

More cases · Software & development

• Update a code base with Copilot chat and agent mode
• Check code against user intent with review agents and tests
• Build a budgeting app from a single prompt
• Build a mental models framework with a multi-agent dev team
• Build a instant-opening markdown viewer for macOS