CtrlK
BlogDocsLog inGet started
Tessl Logo

cookbook-audit

Audit an Anthropic Cookbook notebook based on a rubric. Use whenever a notebook review or audit is requested.

72

1.96x
Quality

56%

Does it follow best practices?

Impact

100%

1.96x

Average score across 3 eval scenarios

SecuritybySnyk

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./.claude/skills/cookbook-audit/SKILL.md
SKILL.md
Quality
Evals
Security

Evaluation results

100%

57%

Cookbook Review: Customer Feedback Analyzer

Audit report format and scoring structure

Criteria
Without context
With context

Executive Summary section

37%

100%

Overall score format

0%

100%

Four scoring dimensions

66%

100%

Each dimension scored /5

0%

100%

Scores sum to 20

0%

100%

Specific Recommendations section

87%

100%

Examples & Suggestions section

50%

100%

Concrete excerpt references

90%

100%

Section justifications

90%

100%

Key Strengths and Critical Issues

16%

100%

100%

74%

Cookbook Review: Image Classification Pipeline

Setup anti-pattern detection

Criteria
Without context
With context

%%capture missing flag

0%

100%

Separate installs flag

100%

100%

os.environ flag

50%

100%

MODEL constant missing flag

0%

100%

Dated model alias flag

0%

100%

Recommended fix for installs

37%

100%

Recommended fix for env loading

0%

100%

Recommended fix for MODEL

0%

100%

Recommended valid model name

0%

100%

Code Quality score

87%

100%

100%

15%

Cookbook Review: Document Summarization Workflow

Narrative and code presentation quality

Criteria
Without context
With context

Machinery-first intro flag

100%

100%

Vague learning objectives flag

100%

100%

Missing pre-code text flag

100%

100%

Missing post-code insight flag

0%

100%

Generic conclusion flag

100%

100%

Problem-first rewrite suggestion

100%

100%

Action-verb LO suggestion

100%

100%

Conclusion improvement suggestion

62%

100%

Narrative Quality score impact

100%

100%

Comment style flag

100%

100%

Repository
anthropics/claude-cookbooks
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

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.