What does this skill do, and when should you use it?
This skill serves users who need guidance on OpenAI products, APIs, or Codex. It prioritizes the OpenAI Developer Docs MCP for authoritative documentation and uses the Codex manual helper first for broad Codex questions. It also owns model selection, API model migrations, and prompt-upgrade guidance. Because it depends on Codex-oriented workflows and OpenAI documentation tools, porting it to other platforms may require adaptation.
Searches and fetches official OpenAI developer documentation; retrieves API specifications when available for reference, schema, parameter, or required-field questions; runs the Codex manual fetch helper and reads its generated manual and outline; provides sourced guidance on model selection, API migration, prompt upgrades, and Codex configuration; restricts fallback web searches to official OpenAI domains.
- A developer needs current official guidance while implementing an OpenAI API feature
- A team needs to choose an API model or plan a model migration
- A user wants to upgrade prompts while keeping changes narrow and behavior-preserving
- A Codex user needs help with skills, plugins, MCP, hooks, configuration, or other Codex behavior
How do you install this skill?
- Core functionality depends on the remote openaiDeveloperDocs MCP and official-network access; reachability, latency, and failure behavior from mainland-China networks are not established by the supplied files.
- No skill-local LICENSE.txt, version identifier, changelog, or skill-level tests are shown, so license compliance, upgrade regression, and key-path reproducibility require additional review.
- The API-key gate, MCP permission scope, and exact data sent to remote documentation services are not specified; credential and request-data governance should be confirmed before use.
- Shell / CLI
- Network access
- Local filesystem
- MCP Server
Node.js
The README states that skills under skills/.system/ are automatically installed in the latest version of Codex. This skill is located at skills/.system/openai-docs/SKILL.md; the supplied source does not provide a separate installation command.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/openai/skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/.system/openai-docs" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- According to the current official docs, how should I choose a model for a multimodal agent workflow?
Ask a direct question about OpenAI documentation, APIs, models, prompts, or Codex, such as: “According to the current official docs, how should I choose a model for a multimodal agent workflow?” For API reference, schema, parameter, or required-field questions, also verify the API specification when the tool is available.
What are this skill's strengths and limitations?
- Prioritizes official OpenAI developer documentation and the Codex manual
- Covers documentation lookup, model selection, migrations, prompt upgrades, and Codex self-knowledge
- Defines an official-domain fallback when the primary documentation tools are unavailable
- Requires the OpenAI Docs MCP, Codex manual helper, and network access
- Some workflows require shell access, Node.js, and local cache files
- The source provides no test suite, platform verification, or actual individual license-file contents
- The README says the repository is deprecated and points to the OpenAI Plugins repository for current examples
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| OpenAI Developer Docs Assistant this page ✓ OpenAI · Official | 55 · Use with care | ★ 28k | 4mo ago | — |
| OpenAI Official Docs Assistant ✓ OpenAI · Official | 52 · Use with care | ★ 129k | 11d ago | Apache-2.0 |
| Azure OpenAI Deployment Assistant ✓ Microsoft · Official | 52 · Use with care | ★ 2.7k | 3mo ago | MIT |
| Azure Skill Creator ✓ Microsoft · Official | 48 · Use with care | ★ 2.7k | 3mo ago | MIT |
| 9Router Chat Skill | 47 · Use with care | ★ 31k | 17d ago | MIT |
How did FollowSkills review this skill?
The skill restricts primary retrieval to the official OpenAI Docs MCP and limits web fallback to official OpenAI domains; API-backed build tasks also require an API-key gate, making its source and permission boundaries reasonably visible. It does not specify MCP permission scope, user confirmation, sensitive-data handling, external side effects, or rollback, and it depends on external services, so 8 points are deducted.
The workflow, fallback order, bounded-uncertainty behavior, and script error handling are clearly described; the scripts also include cache validation, timeouts, and hash checks. Static review cannot reproduce MCP, network, or script key paths, and no committed test suite or CI coverage is supplied, so the score is kept conservative.
The audience and scenarios cover OpenAI APIs, Codex self-knowledge, model selection, migrations, and prompt upgrades, with several explicit non-fit boundaries. Chinese-language usability is not demonstrated with examples, and reachability of the core Docs MCP and official domains from mainland China is unverified, so 4 points are deducted.
The documentation is well layered and includes source priority, workflows, boundaries, references, scripts, and dependency configuration; the name broadly matches the capability, and maintenance/update guidance is present. Evidence for a skill-local license, versioning, and changelog is absent, while some troubleshooting still depends on the external environment, so 7 points are deducted.
The skill provides actionable guidance for official documentation retrieval, Codex manual access, model migration, and prompt upgrades, with clear citation and stopping rules. No skill-level tests, representative outputs, or task validation are included, so direct usability and marginal benefit over manual lookup are only partially established; the static cap of 7 is used.
The supplied sources expose detailed procedures, script implementations, and source restrictions, providing partial auditability. There is no independent third-party execution evidence, test result, or corroborating validation record, and remote MCP behavior cannot be verified statically, so the score remains below the static maximum.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →