What problem does it solve?
This Skill eliminates common pitfalls and inefficiencies when integrating OpenAI and Codex models into applications, services, CLIs, or agent workflows, ensuring you follow current best practices for reliability, security, and cost control.
Core Features & Use Cases
- Current Model Guidance: Select the right OpenAI or Codex model for general reasoning, coding, low-latency, or agentic workloads based on up-to-date official recommendations.
- Responses API Best Practices: Build new integrations using the modern Responses API instead of legacy patterns, with support for streaming, tool calling, and stateful interactions.
- Structured Output Validation: Implement schema-validated model outputs to avoid unreliable free-form text parsing in application logic.
- Operational Guardrails: Follow proven patterns for API key security, retry logic with backoff for rate limits and transient errors, and non-secret logging of request metadata.
- Use Case Example: If you are building a coding agent that needs to analyze repository changes, use this Skill to select the gpt-5.3-codex model with appropriate reasoning effort, implement structured output validation for review decisions, and set up secure key handling and retries for long-running analysis tasks.
Quick Start
Use the openai-api skill to build a new user billing review workflow that calls the OpenAI Responses API with structured output validation and secure API key handling.