audit-agents-skills

Comprehensive quality audit for Claude Code agents, skills, and commands with comparative analysis

Install with Tessl CLI

npx tessl i github:FlorianBruniaux/claude-code-ultimate-guide --skill audit-agents-skills

What are skills?

2.08x

Quality

33%

Does it follow best practices?

Impact

98%

2.08x

Average score across 3 eval scenarios

Optimize this skill with Tessl

npx tessl skill review --optimize ./examples/skills/audit-agents-skills/SKILL.md

SKILL.md

Review

Evals

Evaluation results

99%

86%

CI/CD Integration Audit for Claude Code Assets

JSON audit output structure and scoring

Criteria

Without context

With context

Score as percentage

75%

100%

Grade A threshold

100%

Grade B threshold

80%

Grade C/D/F thresholds

100%

80% production threshold

12%

100%

metadata key

100%

summary key

50%

100%

by_type breakdown

100%

files array fields

20%

100%

failed_criteria fields

100%

top_issues structure

100%

Agents max points

100%

Skills max points

100%

Without context: $0.3629 · 1m 45s · 12 turns · 16 in / 7,591 out tokens

With context: $1.2202 · 5m 24s · 28 turns · 5,483 in / 20,841 out tokens

100%

38%

Quality Audit for Shared Claude Code Asset Library

Directory scanning and dual-format report generation

Criteria

Without context

With context

.claude/agents/ scanned

100%

.claude/commands/ scanned

100%

examples/agents/ scanned

100%

examples/skills/ scanned

100%

Markdown report produced

100%

JSON report produced

100%

Markdown summary table

100%

Markdown per-file scores

100%

Markdown recommendations

100%

Agents scored out of 32

100%

Commands scored out of 20

100%

Skills scored out of 32

100%

Full audit criteria count

100%

Without context: $0.3525 · 2m 3s · 10 turns · 15 in / 7,848 out tokens

With context: $1.1941 · 4m 54s · 32 turns · 217 in / 19,313 out tokens

96%

29%

Transparent Audit with Actionable Fix Recommendations

Fix suggestions format and detection methodology

Criteria

Without context

With context

Fix has file path header

75%

100%

Fix includes points_lost

25%

50%

Fix includes content to add

100%

Overlap: Jaccard numeric value

100%

Overlap: 0.5 threshold used

100%

Token count: numeric value

100%

Token count: word × 1.3 formula

100%

Keyword detection documented

100%

Case-insensitive matching noted

50%

100%

insight-reporter Bash justification flag

100%

Without context: $0.5012 · 3m 17s · 11 turns · 16 in / 12,889 out tokens

With context: $1.1758 · 5m 45s · 19 turns · 5,222 in / 24,188 out tokens

Evaluated: 5 days ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

CI/CD Integration Audit for Claude Code Assets Quality Audit for Shared Claude Code Asset Library Transparent Audit with Actionable Fix Recommendations

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.