ralph

Self-referential loop until task completion with architect verification

Quality

56%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./plugins/oh-my-codex/skills/ralph/SKILL.md

Quality

Content

77%

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 with a clear, validated workflow and feedback loops, but it is long and repetitive with no external reference files. Splitting advanced sections into references would improve token efficiency and disclosure.

Suggestions

Move PRD Mode, Goal Mode Integration, and State Management into separate reference files and link to them from a concise overview.

Deduplicate the rules repeated across Execution Policy, Steps, Tool Usage, and Final Checklist to reduce token cost.

Keep only the core loop and verification steps inline; push the Codex/role-routing diagnostics into a reference for users who hit that surface.

Dimension	Reasoning	Score
Conciseness	The ~290-line body is dense and restates overlapping rules across Execution Policy, Steps, Tool Usage, Goal Mode, State Management, and Final Checklist, so it could be tightened.	2 / 3
Actionability	It provides fully executable commands, concrete JSON state shapes, and copy-paste-ready delegation calls rather than pseudocode.	3 / 3
Workflow Clarity	Steps 0-9 are clearly sequenced with explicit validation checkpoints, fix/re-verify and rollback feedback loops, stop conditions, and a final checklist.	3 / 3
Progressive Disclosure	No bundle files exist and all content is inline in one monolithic file; sections are well-labeled but PRD/Goal/State sections could be split into one-level-deep references.	2 / 3
	Total	10 / 12 Passed

Description

35%

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 captures the core idea (loop until verified) but is abstract, jargon-heavy, and lacks an explicit trigger clause. It would benefit from natural user phrasings and a clear "Use when..." sentence.

Suggestions

Add an explicit "Use when..." clause with natural trigger phrases like "don't stop", "finish this", or "keep going until done" that the body already lists.

Replace jargon ("self-referential loop") with concrete user-facing language so the trigger terms match what users actually say.

Name a couple of concrete capabilities or examples to raise specificity beyond a single mechanism.

Dimension	Reasoning	Score
Specificity	"loop until task completion with architect verification" names a concrete mechanism but gives no domain or multiple specific actions, matching the middle anchor.	2 / 3
Completeness	It loosely conveys what it does but has no "Use when..." trigger clause, so per the guidelines completeness is capped at 2.	2 / 3
Trigger Term Quality	Terms like "self-referential loop" and "architect verification" are internal jargon with no natural user phrasings in the description.	1 / 3
Distinctiveness Conflict Risk	The meta-execution-loop niche is fairly distinctive but the description alone lacks explicit triggers that fully separate it from generic persistence skills.	2 / 3
	Total	7 / 12 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: Yeachan-Heo/oh-my-codex
Path: plugins/oh-my-codex/skills/ralph/SKILL.md
Commit: 435d4a9

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.