codex-first

Claude Code work routing: delegate implementation, fixing, exploratory subagents, rebasing, and PR merging/landing to Codex CLI while the parent specifies, decides, reviews, and verifies. Load a private codex-next policy first when available; otherwise use the native-Claude model gate. Codex-backed autoreview is always allowed and preferred.

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Critical

Do not install without reviewing

Fix and improve this skill with Tessl

tessl review fix ./skills/codex-first/SKILL.md

Quality

Content

88%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 highly actionable and clearly sequenced with strong validation checkpoints and feedback loops for a destructive/batch-adjacent operation. Its main weakness is minor repetition across sections and a large monolithic structure with no progressive disclosure into reference files.

Suggestions

De-duplicate the --yolo/--approve-for-me fallback and the codex-next invocation note so each appears once, with the later occurrences cross-referencing the canonical statement.

Consider moving the install/wrapper setup and the full watchdog script into a references/ file, keeping SKILL.md as a lean overview that links out one level deep.

Tighten the two narrative anecdotes (minification incident, currency/percentage helper) to the single sentence that conveys the lesson.

Dimension	Reasoning	Score
Conciseness	Mostly efficient operational tribal knowledge Claude would not already know, but has minor trimmable repetition (the --yolo/--approve-for-me fallback and the codex-next invocation each recur across sections) and a few narrative anecdotes that could be tightened.	4 / 5
Actionability	Provides fully executable, copy-paste-ready commands with pinned flags — the codex exec invocation, the exec-wrapper install script, the curl route diagnostic, the resume command, and the watchdog loop — covering the common cases concretely.	5 / 5
Workflow Clarity	Multi-step processes are clearly sequenced with explicit validation checkpoints and feedback loops: the model-gate decision tree, the CI-wait precheck, the worker-dies diagnosis table, the liveness watchdog with staleness detection and resume recovery, and the coordinator verification checklist.	5 / 5
Progressive Disclosure	Well-organized into clearly headed sections (Hard gate, Route, Invoke, worker-dies, watchdog, Prompt contract, verification) with easy navigation, though it is a single ~270-line monolithic file with no bundle references, and some content (install wrapper, watchdog script) could arguably live in separate reference files.	4 / 5
	Total	18 / 20 Passed

Description

71%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 specific and distinct, listing concrete delegation and retention actions, but it lacks an explicit 'Use when...' trigger clause, which caps completeness. Some internal jargon slightly weakens trigger-term naturalness.

Suggestions

Add an explicit 'Use when...' clause stating when to invoke this skill (e.g., 'Use when hands-on coding, fixing, rebasing, or PR landing work should move to Codex CLI').

Trim internal jargon like 'codex-next policy' and 'native-Claude model gate' from the description, or pair them with plain-language synonyms.

Add a couple of natural trigger synonyms (e.g., 'land PRs', 'delegate coding to Codex') to broaden keyword coverage.

Dimension	Reasoning	Score
Specificity	Names multiple concrete routing actions — 'delegate implementation, fixing, exploratory subagents, rebasing, and PR merging/landing to Codex CLI while the parent specifies, decides, reviews, and verifies' — giving comprehensive coverage of what the skill routes and what it retains.	5 / 5
Completeness	The 'what' is explicit and clear, but there is no 'Use when...' clause or equivalent explicit trigger guidance — the 'when' is only weakly implied, which per the rubric caps completeness at 3.	3 / 5
Trigger Term Quality	Includes natural terms a user of this workflow would say ('delegate', 'rebasing', 'PR merging/landing', 'Codex CLI', 'autoreview') but mixes in internal jargon ('codex-next policy', 'native-Claude model gate') and omits a few common synonyms, so a couple of natural terms are missing.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (Codex CLI delegation routing for Claude Code) with distinct triggers, but has minor overlap risk with general orchestrator/delegation skills it itself references ('$maintainer-orchestrator').	4 / 5
	Total	16 / 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
referenced_paths_exist	Referenced path issues: 3 missing	Warning

	Total	15 / 16 Passed

Repository: steipete/agent-scripts
Path: skills/codex-first/SKILL.md
Commit: efa3b2f

Reviewed: ~7 hours 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.