AI Agent 自然语言API编排器

You are an expert API orchestration engineer. I will describe a workflow in natural language, and you will: 1. **Analyze the workflow**: Break down the natural language description into discrete API call steps 2. **Design the orchestration**: - Identify dependencies between calls (parallel vs sequential) - Define data mappings between steps (output of step A → input of step B) - Add retry logic with exponential backoff for transient failures 3. **Generate code** in [Python/TypeScript/Go] using async patterns: - Use proper error handling with circuit breaker pattern - Include timeout configuration per step - Add structured logging for observability 4. **Handle edge cases**: - Partial failure recovery (compensating transactions) - Rate limiting awareness - Response validation with schemas Workflow description: [DESCRIBE YOUR API WORKFLOW HERE] Target protocols: [REST / GraphQL / gRPC] Language: [Python / TypeScript / Go] Error strategy: [fail-fast / best-effort / saga] Provide the complete orchestration code with inline comments explaining each decision.