openai-api-core

openai-api-core

Official reference covering text generation, structured outputs, streaming, function calling, and conversation state management, centered on the OpenAI API — Responses API. Includes webhooks, error codes, rate limits, official SDKs (Python / JS / .NET / Java / Go / Ruby), and migration guides from Chat Completions / Assistants.

Directory structure

skills/openai-api-core/
  SKILL.md
  references/
    getting-started/
      README.md
      quickstart.md
      concepts.md
      models.md
      models-all.md
      models-compare.md
    responses/
      README.md
      text.md
      conversation-state.md
      prompt-caching.md
      reasoning.md
      reasoning-best-practices.md
      background.md
      responses-multi-agent.md
      compaction.md
      deployment-checklist.md
    structured-streaming/
      README.md
      structured-outputs.md
      streaming-responses.md
      function-calling.md
      programmatic-tool-calling.md
      websocket-mode.md
      prompt-generation.md
    webhooks-errors/
      README.md
      webhooks.md
      realtime-server-controls.md
      error-codes.md
      rate-limits.md
    libraries/
      README.md
      python-sdk.md
      javascript-sdk.md
      dotnet-sdk.md
      java-sdk.md
      go-sdk.md
      ruby-sdk.md
      openai-cli.md
    legacy-migration/
      README.md
      responses-vs-chat-completions.md
      migrate-chat-completions-to-responses.md
      migrate-from-prompt-objects.md
      assistants-to-responses-migration.md
      deprecations.md
  samples/
    README.md
    basic-text-generation.md
    structured-outputs.md
    streaming-responses.md
    multi-turn-conversation.md
    function-calling.md
    webhook-verification.md
    background-responses.md
  scripts/
    README.md
    install.md
    auth.md
    curl-endpoints.md
    cli.md

Exploration procedure

Derive the category from the task, and identify the target page in the category's README.md:

1. Find the category corresponding to the task in the mapping table below
2. For that category, references/{category}/README.md（samples/ scripts/ (the README.md directly under it) to identify the target page
3. Read the .md of the relevant page to check the details

For the Agents SDK, built-in tools (web search / file search / code interpreter / computer use), and MCP integration, refer to the openai-agents skills, Codex (coding agent CLI) is openai-codex skill. This skill handles how to call the Responses API itself, structured outputs, streaming, webhooks, official SDKs, and legacy migration.

Task → Category mapping

Task	Category	Reference README
I want to know about API key setup, SDK installation, and making the first API call	getting-started	references/getting-started/README.md
I want to know the model list, selection criteria, and comparison of context windows / max output	getting-started	references/getting-started/README.md
I want to generate text with the Responses API and learn the basics of prompt design	responses	references/responses/README.md
I want to handle conversation state management (previous_response_id / Conversations API) and prompt caching	responses	references/responses/README.md
For reasoning models, effort/summary settings, background mode, multi-agent, and compaction	responses	references/responses/README.md
I want to check the design checklist before production deployment (deployment checklist)	responses	references/responses/README.md
I want to obtain structured outputs (JSON Schema, strict mode)	structured-streaming	references/structured-streaming/README.md
I want to receive responses incrementally via SSE streaming	structured-streaming	references/structured-streaming/README.md
I want to call tools via function calling / programmatic tool calling	structured-streaming	references/structured-streaming/README.md
I want to build an always-connected agentic workflow over WebSocket	structured-streaming	references/structured-streaming/README.md
I want to auto-generate prompts / function definitions / schemas with the Playground Generate feature	structured-streaming	references/structured-streaming/README.md
I want to know about receiving and verifying webhooks and server-side control of the Realtime API	webhooks-errors	references/webhooks-errors/README.md
I want to know about error codes, exception classes, rate limits, and backoff strategies	webhooks-errors	references/webhooks-errors/README.md
I want to know how to use the official SDKs (Python / JavaScript / .NET / Java / Go / Ruby)	libraries	references/libraries/README.md
I want to know about installing and using commands for the openai CLI	libraries	references/libraries/README.md
I want to decide whether to use the Responses API or the Chat Completions API	legacy-migration	references/legacy-migration/README.md
I want to migrate from Chat Completions / Assistants to the Responses API and know the deprecation schedule	legacy-migration	references/legacy-migration/README.md
I want to migrate from managed prompt objects to in-code version management	legacy-migration	references/legacy-migration/README.md
I want to know typical usage (text generation, structured outputs, streaming, conversation continuation, function calling, webhook verification, background execution, etc.)	samples	samples/README.md
I want to know about installation, authentication settings, calling endpoints with curl, and CLI commands	scripts	scripts/README.md