ralph-skill-review-loop

Self-improving review loop for Ralph Wiggum skills. Reviews skills against best practices, implements improvements, and continues until two consecutive clean reviews. Use when validating or improving the ralph-prompt-* skill suite.

Quality

66%

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/ralph-skill-review-loop/SKILL.md

Quality

Content

62%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 a thorough, well-sequenced review loop with strong workflow clarity and actionable commands. Its weaknesses are verbosity from duplicated criteria and padded meta-sections, plus a monolithic structure that forgoes progressive disclosure via separate reference files.

Suggestions

Remove the duplicated review-criteria list (REVIEW CRITERIA REFERENCE) since the same 16 criteria already appear in STEP 2, and consolidate the repeated RALPH_REVIEW_STATE.json snippets into one canonical example.

Extract the large embedded review-loop prompt template into a separate reference file (e.g. references/review-loop-prompt.md) and keep SKILL.md as a concise overview that links to it, improving progressive disclosure.

Trim non-actionable sections such as 'Why This Works', 'Expected Behavior', and 'Customization' to reduce token cost while preserving the executable workflow.

Dimension	Reasoning	Score
Conciseness	At ~530 lines the body is noticeably verbose: the 16 review criteria are listed twice (STEP 2 and REVIEW CRITERIA REFERENCE), state-JSON snippets are repeated across multiple steps, and sections like 'Why This Works', 'Expected Behavior', and 'Customization' add padding without new actionable information.	2 / 5
Actionability	Provides concrete, runnable bash/git/jq commands and explicit checklists, but the many bracketed placeholders ([N], [count], [brief description]) throughout the templates leave minor gaps versus copy-paste-ready material.	4 / 5
Workflow Clarity	The six-step loop is clearly sequenced with explicit validation (STEP 4 post-improvement verification), feedback loops (review -> fix -> re-review), termination conditions, checklists, and an escape hatch, matching the top anchor.	5 / 5
Progressive Disclosure	Section headers organize the content, but the skill is a monolithic single file with the large embedded prompt template and reference criteria inlined; no bundle files exist, and content that would benefit from being split into separate references is kept inline.	3 / 5
	Total	14 / 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 clearly conveys purpose and trigger context in a concise, third-person voice with a distinct niche. Its main limitation is trigger-term breadth, which offers only a couple of natural keywords without synonyms.

Dimension	Reasoning	Score
Specificity	Names several concrete actions ('Reviews skills against best practices, implements improvements, and continues until two consecutive clean reviews') with only minor coverage gaps, matching the 'several specific actions' anchor rather than the comprehensive 5.	4 / 5
Completeness	Explicitly states both what it does (review, improve, loop until two clean reviews) and when to use it ('Use when validating or improving the ralph-prompt-* skill suite'); the when clause is present but could be more specific, so it sits below the 5 anchor.	4 / 5
Trigger Term Quality	Includes natural terms 'validating' and 'improving' tied to the ralph-prompt-* suite, but coverage is thin with missing common synonyms or variations, fitting the 'some relevant keywords but missing variations' anchor.	3 / 5
Distinctiveness Conflict Risk	Targets a clearly defined niche ('ralph-prompt-* skill suite') with distinct triggers and minimal overlap risk with other skills.	5 / 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
skill_md_line_count	SKILL.md is long (539 lines); consider splitting into references/ and linking	Warning

	Total	15 / 16 Passed

Repository: fernandezbaptiste/Skrillz
Path: skills/ralph-skill-review-loop/SKILL.md
Commit: 93ed392

Reviewed: 1 day 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.