CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

openai-docs

Use for Codex models/pricing, scheduled tasks, skills, settings, setup, troubleshooting, customization, automations, and self-knowledge—including 'you,' 'your,' 'this app,' or 'this coding agent' when they refer to Codex—and for OpenAI APIs/products and ChatGPT Work. Also use for model choice/migration, prompting, SDKs, Responses, Realtime, agents, evals, and Chat/Work/Codex comparisons. Do not use for generic app/software tasks that merely mention Codex.

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

78%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A disciplined, routing-style skill body: it mandates a search-first source order, defines five condition-gated routes with a one-reference cap, and closes with concrete citation, domain, and disclosure boundaries. Main weaknesses are redundancy in the opening paragraph and rule restatements, plus no final verification checkpoint before answering.

Suggestions

State the preserve-model rule once in 'Source and execution boundaries' and remove the duplicate phrasings in the first-paragraph and model-migration-route lines.

Tighten the 'First substantive action' paragraph: the search-then-actually-open instruction is repeated three ways; state it once and keep the exception order as a short list.

Add an explicit closing checkpoint such as 'Before answering, confirm the fetched page supports each pricing/availability/limits claim; otherwise state uncertainty' to complete the workflow with a verification step.

DimensionReasoningScore

Conciseness

No known-concept padding and the body is dense and instructional, but rules repeat: "Preserve the exact requested model; never substitute a newer model" is restated as "Preserve an explicitly requested target" and again as "Preserve an explicitly requested model for selection, migration, and prompting", and "actually open or fetch" appears three times in the first paragraph.

4 / 5

Actionability

Concrete, executable directives throughout: "Search the user's exact requested... topic... using a concise, topic-specific query of 2-6 essential terms", "Search, open, fetch, and cite only `developers.openai.com`, `platform.openai.com`, and `learn.chatgpt.com`", "use `openai-platform-api-key` first", "Say 'OpenAI Docs' or 'official OpenAI documentation' in user-facing answers".

4 / 5

Workflow Clarity

Sequence is explicit ("Complete this source order before reading a reference, inspecting local or repository files... or answering from memory"), routes are prioritized ("Use the first matching route"), and a feedback loop exists ("If one official search or page does not establish the answer, search another appropriate official domain").

4 / 5

Progressive Disclosure

The body is a true overview ("Read zero or one primary reference", "Read at most one primary reference") where every route signals a real one-level-deep file with an explicit only-when condition; all 8 referenced files exist in references/, and every bundle file is reachable (upgrade-guide.md is linked from model-migration.md).

5 / 5

Total

17

/

20

Passed

Description

86%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, dense description with explicit use/skip triggers, comprehensive natural keywords including synonyms and Codex pronoun references, and a clear niche protected by a do-not-use clause. Its only gap is that the capability ('what it does') is implied by topic lists rather than stated as an action.

DimensionReasoningScore

Specificity

Names many specific areas — "Codex models/pricing, scheduled tasks, skills, settings, setup, troubleshooting, customization, automations... OpenAI APIs/products and ChatGPT Work... SDKs, Responses, Realtime, agents, evals" — but describes scope rather than concrete actions; no capability verb (e.g., 'provides' or 'answers with') states what the skill does.

4 / 5

Completeness

'When' is explicit and detailed ("Use for... Also use for... Do not use for generic app/software tasks that merely mention Codex") with concrete trigger phrases, but 'what' is only implied by the topic enumeration — the description never states the action performed (current, cited OpenAI documentation guidance).

4 / 5

Trigger Term Quality

Comprehensive natural-term coverage with synonyms and product variations users actually say: "model choice/migration", "prompting", "SDKs", "Responses", "Realtime", "evals", "Chat/Work/Codex comparisons", plus pronoun triggers "'you,' 'your,' 'this app,' or 'this coding agent'".

5 / 5

Distinctiveness Conflict Risk

Clear niche (official OpenAI/Codex documentation and self-knowledge) with an explicit anti-trigger — "Do not use for generic app/software tasks that merely mention Codex" — that preempts the most likely false activation, keeping conflict risk minimal.

5 / 5

Total

18

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

Total

15

/

16

Passed

Repository

openai/codex

Path

codex-rs/skills/src/assets/samples/openai-docs/SKILL.md

Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.