ai-md

Convert human-written CLAUDE.md into AI-native structured-label format. Battle-tested across 4 models. Same rules, fewer tokens, higher compliance.

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/ai-md/SKILL.md

Quality

Content

63%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.

This is a well-structured progressive-disclosure skill: a lean, navigable body pointing to a single self-contained detailed guide that carries the full validated workflow. Its main weakness is that the body provides no inline quick-start or executable snippet, deferring all actionability to the bundle.

Suggestions

Add a short inline 'Quick start' snippet in the body (e.g. the PREVIEW token-measurement bash block or a 3-line distilled-template skeleton) so Claude has immediate executable guidance without first loading the guide.

Surface the core validation checkpoint at the top level (e.g. 'Mandatory: back up CLAUDE.md and run the 2+ model, 8-question test before keeping any conversion') so the destructive-safety gate is visible without loading the bundle.

Trim the 'Detailed Guide' paragraph to one sentence (e.g. 'Read references/detailed-guide.md before executing; its safety and validation requirements are mandatory.') to push conciseness toward 5.

Dimension	Reasoning	Score
Conciseness	The 25-line body is lean, assumes Claude's competence, and avoids explaining concepts Claude knows; the only trimmable padding is the 'Detailed Guide' paragraph ('It retains the complete procedure and reference material'), which keeps it just below the 5 anchor.	4 / 5
Actionability	The body itself contains no executable code or commands and only defers execution to the guide ('Read the detailed guide before executing this skill'), matching the 'minimal concrete guidance; high-level hints but missing the specific steps' anchor; the rich actionable material lives entirely in references/detailed-guide.md rather than the body.	2 / 5
Workflow Clarity	The bundle's Two-Stage Workflow (PREVIEW → DISTILL) is clearly sequenced with explicit validation checkpoints (backup, Phase 6 multi-model test, revert-if-score-drops, restore command), and the body signals these as mandatory; it stops short of 5 because the sequence and validation are not surfaced in the body itself.	4 / 5
Progressive Disclosure	The body is a clean overview (Detailed Guide, When to Use, Limitations) with a well-signaled, verified one-level-deep reference to references/detailed-guide.md, which is self-contained with no nested file hops, matching the clear-overview anchor.	5 / 5
	Total	15 / 20 Passed

Description

53%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 communicates what the skill does and occupies a distinct niche, but it omits any 'when to use' trigger guidance and relies on a single concrete action plus benefit claims. Adding an explicit 'Use when...' clause with natural trigger terms would raise both completeness and trigger_term_quality.

Suggestions

Append a 'Use when...' clause with natural triggers, e.g. 'Use when your CLAUDE.md is long but the AI ignores rules, token usage is too high, or you are migrating rules between Claude, Codex, Gemini, and Grok.'

Add one or two more concrete actions (e.g. 'decomposes prose rules into atomic labeled gates, validates across 2+ models') to move specificity from a single action toward comprehensive coverage.

Include user-natural synonyms such as 'system prompt' and 'instructions' alongside 'CLAUDE.md' to improve trigger-term coverage.

Dimension	Reasoning	Score
Specificity	Names the domain ("human-written CLAUDE.md") and one concrete action ("Convert ... into AI-native structured-label format"), but the remainder ("Same rules, fewer tokens, higher compliance") states outcomes rather than additional actions, so it does not reach the 'several specific actions' bar of 4.	3 / 5
Completeness	The 'what' is clear (convert CLAUDE.md to structured-label format) but there is no 'Use when...' or equivalent trigger clause, so per the rubric guideline completeness is capped at 3.	3 / 5
Trigger Term Quality	Relevant keywords appear ("CLAUDE.md", "tokens", "compliance"), but natural user phrasings like "system prompt", "instructions", or "AI ignores my rules" are absent, leaving common synonyms missing as the 3 anchor describes.	3 / 5
Distinctiveness Conflict Risk	"Convert human-written CLAUDE.md into AI-native structured-label format" carves a clear, narrow niche with low overlap risk; it stops short of 5 only because the description lacks explicit distinct trigger phrases.	4 / 5
	Total	13 / 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: sickn33/agentic-awesome-skills
Path: skills/ai-md/SKILL.md
Commit: b1aebac

Reviewed: ~18 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.