CtrlK
BlogDocsLog inGet started
Tessl Logo

claude-md

Create or update CLAUDE.md files following best practices for optimal AI agent onboarding

56

Quality

63%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./03-skills/claude-md/SKILL.md
SKILL.md
Quality
Evals
Security

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 a well-structured, actionable instruction skill with a copy-paste-ready template, clearly sequenced workflows, and an explicit validation checklist. Its main weaknesses are redundancy across the Quality Constraints, Anti-Patterns, and Validation Checklist sections and the absence of progressive disclosure applied to its own content.

Suggestions

Consolidate the overlapping "Quality Constraints", "Anti-Patterns to Avoid", and "Validation Checklist" sections to remove repeated points (no style rules, no code snippets, no redundancy).

Trim concept-explanation framing such as "LLMs are stateless..." that Claude already knows; keep only skill-specific rationale.

Move the detailed "Essential Sections" template and/or anti-patterns into a referenced bundle file (e.g., references/template.md) to apply progressive disclosure to the skill's own ~208-line body.

DimensionReasoningScore

Conciseness

Points recur across sections — "No Style Rules" / "No style/formatting rules" and "No Code Snippets" / "No code snippets (use file references)" appear in both "Quality Constraints" and the "Validation Checklist", and "Less is More" / "less is more" repeats — while "LLMs are stateless: CLAUDE.md is the only file automatically included in every conversation" explains a concept Claude already knows. Mostly efficient and well-organized but could be tightened. Not 3 due to cross-section duplication; not 1 because it is organized, actionable instruction rather than padding.

2 / 3

Actionability

The "Essential Sections" block provides a copy-paste-ready markdown template, concrete file paths ("./CLAUDE.md", ".claude/CLAUDE.md", "~/.claude/CLAUDE.md"), a concrete agent_docs/ structure, and step-by-step Output Format for create/update/audit. Specific examples and copy-paste-ready guidance; per the scoring notes, absence of code in an instruction-only skill is not penalized when guidance is actionable. Not 2 because the guidance is concrete and complete rather than pseudocode or vague.

3 / 3

Workflow Clarity

"Execution Flow" gives numbered steps 1-7, the "Validation Checklist" includes "Before finalizing, verify:" with "Commands are verified to work", and the Output Format spells out separate create/update/audit sequences. Clear sequence with explicit validation steps and a checklist. Not 2 because checkpoints are explicit rather than implicit; this is a non-destructive documentation task so the strict feedback-loop cap does not apply.

3 / 3

Progressive Disclosure

The skill is a single ~208-line SKILL.md with no bundle files (references/scripts/assets absent); it teaches progressive disclosure to the user ("recommend creating an agent_docs/ folder") but applies none to its own content. Well-organized sections, but content that could be split — e.g., the "Essential Sections" template and detailed anti-patterns — is all inline. Not 3 because it exceeds the simple-skill threshold with no one-level-deep references of its own; not 1 because it is well-organized rather than a monolithic wall or nested references.

2 / 3

Total

10

/

12

Passed

Description

50%

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 states what the skill does (create or update CLAUDE.md files) but lacks an explicit "Use when..." trigger and leans on generic "best practices / onboarding" framing, leaving every dimension at the mid level.

Suggestions

Add an explicit "Use when..." trigger clause (e.g., "Use when creating, updating, or auditing a CLAUDE.md for a project") to raise completeness and trigger_term_quality.

Replace the vague "following best practices for optimal AI agent onboarding" with concrete actions or outcomes (e.g., "applying WHAT/WHY/HOW structure, length constraints, and progressive disclosure").

Include natural keyword variations users might say (e.g., "CLAUDE.md", "agent onboarding file", "project context file") to broaden trigger coverage.

DimensionReasoningScore

Specificity

"Create or update CLAUDE.md files" names a concrete domain and two actions, matching the anchor that names a domain and some actions without being comprehensive. Not 3 because "following best practices for optimal AI agent onboarding" is abstract filler rather than additional concrete actions; not 1 because it does specify concrete actions on a concrete file type.

2 / 3

Completeness

"Create or update CLAUDE.md files" answers "what" but there is no "Use when..." clause or explicit trigger for "when". Per the judging guidelines, a missing explicit trigger guidance caps completeness at 2. Not 3 because "when" is absent; not 1 because "what" is clearly stated.

2 / 3

Trigger Term Quality

"CLAUDE.md files" is a natural term a user would actually say, giving some relevant keywords. Not 3 because coverage is narrow — a single core keyword with no common variations and no trigger phrasing; not 1 because "CLAUDE.md" is a genuine natural keyword rather than technical jargon.

2 / 3

Distinctiveness Conflict Risk

"CLAUDE.md files" targets a specific, distinctive niche, but "following best practices for optimal AI agent onboarding" is generic framing and, without explicit distinct triggers, could overlap with broader documentation/onboarding skills. Not 3 because no explicit distinct triggers; not 1 because CLAUDE.md is a concrete, specific target unlikely to broadly conflict.

2 / 3

Total

8

/

12

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
luongnv89/claude-howto
Reviewed

Table of Contents

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.