codex-claude-cursor-loop

Orchestrates a triple-AI engineering loop where Claude plans, Codex validates logic and reviews code, and Cursor implements, with continuous feedback for optimal code quality

Quality

58%

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 ./skills/codex-claude-cursor-loop/SKILL.md

Quality

Content

58%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 skill delivers an executable, well-sequenced multi-agent workflow with strong validation loops, but it is padded with redundant sections and does not use any progressive disclosure via bundle files, hurting conciseness and structure.

Suggestions

Remove redundant sections — 'Recovery When Issues Are Found', 'The Perfect Loop', and 'Best Practices' largely restate the phases and inline notes; consolidate them into the phase flow.

Move the 'Command Reference' table and detailed recovery procedures into reference files under references/ and link to them from the body to introduce one-level-deep progressive disclosure.

Trim repeated injunctions (e.g., repeated 'IMPORTANT: use --resume with same session ID') to a single canonical statement to improve token efficiency.

Dimension	Reasoning	Score
Conciseness	The ~210-line body is noticeably padded: 'Recovery When Issues Are Found' restates Phases 3/5/7, 'Best Practices' repeats the inline session notes, 'The Perfect Loop' ASCII diagram re-sequences the phases, and 'Command Reference' duplicates every command already shown — fitting 'noticeably verbose; several unnecessary padded sections' below the midpoint.	2 / 5
Actionability	It provides concrete, mostly copy-paste-ready commands with real flags ('codex exec -m <model> --config model_reasoning_effort=...', 'cursor-agent --resume=... --sandbox read-only') and explicit session handling, with only minor placeholders like '[Claude's plan here]' keeping it from a 5.	4 / 5
Workflow Clarity	The 7-phase sequence is clearly ordered with explicit validation checkpoints and full feedback loops (Phase 3 plan refinement, Phase 7 iterative fix loop with 'repeat until all validations pass'), matching the 'clear sequence with explicit validation steps and feedback loops' anchor.	5 / 5
Progressive Disclosure	No bundle files exist (references/scripts/assets directories are absent) and everything is inlined into one monolithic SKILL.md with no one-level-deep references or navigation; content that belongs in separate files (command reference, recovery procedures) is inlined, fitting 'minimal structure; content that belongs in separate files is inlined'.	2 / 5
	Total	13 / 20 Passed

Description

58%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 conveys a concrete, distinctive workflow but omits an explicit 'Use when' trigger clause and lacks natural user-facing trigger terms, capping completeness and trigger quality at the midpoint.

Suggestions

Add an explicit 'Use when...' clause naming concrete triggers (e.g., 'Use when building features that need cross-model validation, code review, or multi-agent implementation').

Include natural trigger phrases and synonyms users would say, such as 'code review', 'multi-agent coding', or 'CI-style validation loop'.

Tighten toward comprehensiveness by listing the full set of supported actions (plan, validate, implement, review, iterate) to lift specificity toward 5.

Dimension	Reasoning	Score
Specificity	Names the domain (triple-AI loop) and several concrete per-agent actions — 'Claude plans', 'Codex validates logic and reviews code', 'Cursor implements' — with only minor coverage gaps, fitting the 'lists several specific actions' anchor rather than the comprehensive score 5.	4 / 5
Completeness	It gives a clear 'what' (orchestrates a three-AI validation loop) but has no 'Use when...' or equivalent trigger clause; per the judging guidelines a missing explicit 'when' caps completeness at 3.	3 / 5
Trigger Term Quality	It contains some relevant keywords ('engineering loop', 'code quality', 'plan', 'review') but lacks the natural phrases and synonyms a user would actually say and includes no file extensions, matching 'some relevant keywords but missing common variations'.	3 / 5
Distinctiveness Conflict Risk	The triple-AI orchestration niche is fairly distinct with minimal overlap risk against typical single-purpose skills, fitting 'mostly distinct; minor overlap risk' rather than the fully clear-niche score 5.	4 / 5
	Total	14 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: bear2u/my-skills
Path: skills/codex-claude-cursor-loop/SKILL.md
Commit: 8ed5784

Reviewed: 10 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.