codex

Use when the user asks to run Codex CLI (codex exec, codex resume) or references OpenAI Codex for code analysis, refactoring, or automated editing. Uses GPT-5.2 by default for state-of-the-art software engineering.

Quality

77%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./codex/codex/SKILL.md

Quality

Content

67%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.

The body is actionable and well-structured with concrete commands and validation prompts, but is held back by a duplicate numbered step, a contradictions in the resume flag guidance, and time-sensitive/marketing content that bloats the token budget.

Suggestions

Fix the duplicate step numbering: two items are both labeled '3.' (the flags list and 'Always use --skip-git-repo-check') — renumber sequentially.

Reconcile the resume-flag guidance: step 4 says flags can be inserted between exec and resume while the quick-reference table says 'no flags allowed' — state one rule clearly.

Move version-specific figures (SWE-bench percentages, v0.57.0, knowledge cutoff) and benchmark/marketing claims into a 'Version / Notes' section so the core workflow stays lean and doesn't decay.

Correct the typo 'if he species the model' to 'if the user specifies the model'.

Dimension	Reasoning	Score
Conciseness	Mostly efficient command/flag lists, but time-sensitive figures (SWE-bench 76.3% vs 72.8%, v0.57.0, knowledge cutoff) and marketing claims ('reduced hallucinations', '30% faster') sit in the main content rather than a deprecated/version section, adding non-essential tokens.	3 / 5
Actionability	Provides copy-paste-ready commands including the exact resume one-liner and a quick-reference table mapping use cases to flags; minor gaps come from the slightly contradictory 'no flags allowed' vs 'insert flags between exec and resume' wording.	4 / 5
Workflow Clarity	A clear 7-step sequence with checkpoints (AskUserQuestion for effort level and high-impact flags, stop-on-non-zero error handling), but the duplicate '3.' step numbering and the resume-flags contradiction are real sequencing flaws.	4 / 5
Progressive Disclosure	Single well-sectioned SKILL.md with no nested references and clean headers (Running a Task, Safety Boundaries, Quick Reference, Model Options); the inline model-options/marketing table is the only content that might warrant a separate reference file.	4 / 5
	Total	15 / 20 Passed

Description

87%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.

The description is well-formed with an explicit trigger clause, concrete actions, and distinct command-level keywords that minimize mis-trigger risk. Minor trim opportunity is the marketing-flavored final sentence.

Suggestions

Drop or relocate 'Uses GPT-5.2 by default for state-of-the-art software engineering' — the model default is an implementation detail already covered in the body and 'state-of-the-art' is unverifiable fluff.

Add a couple of natural synonyms (e.g. 'code review', 'code generation') to broaden trigger coverage without losing distinctiveness.

Dimension	Reasoning	Score
Specificity	Names the domain (Codex CLI) with sub-commands (codex exec, codex resume) and several concrete actions (code analysis, refactoring, automated editing), but coverage is not fully comprehensive of all Codex capabilities.	4 / 5
Completeness	Explicitly answers both what (run Codex CLI for code analysis/refactoring/automated editing, defaults to GPT-5.2) and when via a concrete 'Use when the user asks...' trigger clause.	5 / 5
Trigger Term Quality	Includes natural command-level terms ('codex exec', 'codex resume', 'OpenAI Codex') and task phrases users would say, though a few synonyms or phrasings are absent.	4 / 5
Distinctiveness Conflict Risk	The Codex-specific command triggers (codex exec, codex resume, OpenAI Codex) carve a clear niche with minimal overlap risk; the trailing 'state-of-the-art' clause is generic but does not broaden the trigger.	5 / 5
	Total	18 / 20 Passed

Validation

93%

Warnings & errors only

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

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: jdrhyne/agent-skills
Path: codex/codex/SKILL.md
Commit: 439cd3a

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.