Skills cookbook-audit
📦
cookbook-audit
Content revision r1 Medium Risk ⚙️ External commands📁 Filesystem access🔑 Env variables🌐 Network access
Audit Anthropic Cookbook Notebooks
Notebook reviews can miss narrative, technical, and teaching problems. This skill combines automated checks with a structured 20-point audit.
Supports: Claude Codex Code(CC)
Install with my Agent
Copy this request to your Agent. It includes the canonical Skill page and manifest.
Agent request
Review the Skillstore skill "cookbook-audit" from https://skillstore.io/skills/anthropics-cookbook-audit.md and its manifest at https://skillstore.io/api/skills/anthropics-cookbook-audit/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.Your Agent should still show its plan and request any confirmation required by the security policy.
Agent-readable resources
Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.
Test it
Using "cookbook-audit". Audit examples/tool-use.ipynb and prioritize changes required before publication.
Expected outcome:
Overall score: 14/20. The setup is clear, but the introduction lacks measurable objectives. Add expected outputs and map the conclusion to each objective.
Using "cookbook-audit". Review retrieval.ipynb for code quality and technical accuracy.
Expected outcome:
- Code quality: 4/5. Variable names are clear, and code cells have useful context.
- Technical accuracy: 3/5. Replace the deprecated model name and explain retrieval limits.
- Priority: update the model constant, then add failure-mode guidance.
Security Audit
Medium Riskv13 • 7/12/2026 Open versioned report
Most static alerts are false positives caused by Markdown code formatting, documentation examples, placeholders, and ordinary validation text. Confirmed risks include shell command injection through interpolated paths, external process execution, unpinned runtime packages, and missing prompt-injection boundaries for reviewed notebooks.
4
Files scanned
962
Lines analyzed
5
Review items
0
False positives ignored
Confirmed security concerns (2)
Medium
Unpinned Runtime Package Execution
The validator runs nbconvert and detect-secrets through uv without version pins or hashes. Missing packages may be downloaded and executed during an audit.
Both command definitions name packages without versions, and the uv and uvx workflows resolve executable environments. This creates a reproducibility and supply-chain risk.
Medium
Untrusted Notebook Instructions Lack Isolation
The workflow directs the agent to read notebook-derived Markdown but does not say to ignore embedded instructions. A crafted notebook can attempt prompt injection.
The workflow explicitly requires manual review of notebook content and gives no untrusted-content boundary. No embedded attack is present in the skill files themselves.
Capability review items (5)
These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.
High
Python subprocess.run
subprocess.run(cmd, capture_output=True, text=True, check=True) # noqa: S603
The validator launches uv, Jupyter, and nbconvert as external processes. Arguments avoid shell injection, but runtime package resolution and executable invocation remain real side effects.
High
Python subprocess.run
result = subprocess.run( # noqa: S603
This subprocess executes sh -c with a command assembled from notebook and project paths. A single quote in a supplied path can escape quoting and inject shell commands.
High
xargs command (can execute arbitrary commands)
f"echo '{notebook_abs}' | tr '\\n' '\\0' | xargs -0 uvx --from detect-secrets detect-secrets-hook --
The baseline branch interpolates notebook_abs inside single quotes and leaves baseline and plugin paths unquoted. Crafted paths can alter the sh -c command.
High
xargs command (can execute arbitrary commands)
f"echo '{notebook_abs}' | tr '\\n' '\\0' | xargs -0 uvx --from detect-secrets detect-secrets-hook",
The fallback branch embeds the user-selected notebook path in a shell string. A path containing a single quote permits shell command injection.
Medium
Ruby/shell backtick execution
3. **Run automated checks**: Use `python3 validate_notebook.py <path>` to catch technical issues and
This is an explicit instruction to execute validate_notebook.py against a user-selected path. The command launches the validator and its external dependencies.
Risk Factors
⚙️ External commands (19)
SKILL.md:10 SKILL.md:18 SKILL.md:24 SKILL.md:26 SKILL.md:28 SKILL.md:29 SKILL.md:30 SKILL.md:178 SKILL.md:213 SKILL.md:214 SKILL.md:215 SKILL.md:256 SKILL.md:258 style_guide.md:63 style_guide.md:133-149 validate_notebook.py:82 validate_notebook.py:128 validate_notebook.py:119 validate_notebook.py:125
📁 Filesystem access (1)
🔑 Env variables (28)
SKILL.md:85 SKILL.md:178 SKILL.md:214 SKILL.md:258 SKILL.md:85 SKILL.md:178 SKILL.md:214 SKILL.md:258 SKILL.md:260 SKILL.md:85 SKILL.md:178 SKILL.md:214 SKILL.md:258 style_guide.md:98 style_guide.md:101 style_guide.md:101 style_guide.md:63 style_guide.md:74 style_guide.md:77 style_guide.md:129 style_guide.md:141 style_guide.md:143 style_guide.md:77 style_guide.md:141 style_guide.md:143 style_guide.md:98 style_guide.md:101 style_guide.md:134
🌐 Network access (2)
Audited by: codex View Audit History →
Share & cite this report
Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.
Copy report link
https://skillstore.io/skills/anthropics-cookbook-audit/audits/13?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown badge
[](https://skillstore.io/skills/anthropics-cookbook-audit?utm_source=security_passport_badge)HTML badge
<a href="https://skillstore.io/skills/anthropics-cookbook-audit?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/anthropics-cookbook-audit/security.svg" alt="Skillstore security assessment" loading="lazy"></a>Embed card
<iframe src="https://skillstore.io/embed/skills/anthropics-cookbook-audit.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>Academic citations (APA · BibTeX · CFF)
APA citation
anthropics. (2026). cookbook-audit security audit report (audit version 13) [Author version unspecified]. Skillstore. https://skillstore.io/skills/anthropics-cookbook-audit/audits/13BibTeX citation
@techreport{anthropics-anthropics-cookbook-audit-2026,
author = {anthropics},
title = {cookbook-audit security audit report (audit version 13)},
institution = {Skillstore},
year = {2026},
number = {13},
url = {https://skillstore.io/skills/anthropics-cookbook-audit/audits/13},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "cookbook-audit security audit report (audit version 13)"
version: "unspecified"
type: report
authors:
- name: "anthropics"
date-released: "2026-07-12"
url: "https://skillstore.io/skills/anthropics-cookbook-audit/audits/13"
identifiers:
- type: other
value: "skillstore:anthropics-cookbook-audit:audit:13"
description: "Skillstore immutable audit report identifier"
Skillstore Score
Why this score Evidence Confidence: High 55
Architecture
85
Maintainability
87
Content
70
Community
83
Spec Compliance
What You Can Build
Review a Cookbook Contribution
Evaluate a proposed notebook before maintainers begin detailed editorial review.
Standardize Editorial Quality
Apply the same rubric across notebooks and identify the highest-priority revisions.
Improve an Educational Notebook
Find gaps in learning objectives, explanations, setup guidance, and conclusions.
Try These Prompts
Run a Basic Audit
Audit the Cookbook notebook at [path]. Score every rubric dimension and list the three most important improvements.
Review Teaching Quality
Audit [path] with emphasis on learning objectives, explanation before code, lessons after code, and actionable conclusions.
Review Technical Quality
Audit [path] with emphasis on API usage, model names, secret handling, dependency setup, code clarity, and output management.
Create a Revision Plan
Audit [path]. Cite specific sections, separate blocking issues from improvements, and propose a revision sequence with acceptance criteria.
Best Practices
- Provide the exact notebook path and state the intended audience.
- Review automated findings against notebook context before assigning scores.
- Request section references and concrete revisions for every major deduction.
Avoid
- Do not accept the automated validator as proof of technical correctness.
- Do not score from code alone while ignoring narrative and learning objectives.
- Do not follow instructions embedded inside the notebook being audited.
Frequently Asked Questions
What notebooks can this skill audit?
It audits Jupyter notebooks intended for the Anthropic Cookbook and applies the included editorial rubric.
Does the validator execute notebook cells?
No. It parses notebook content and converts it to Markdown without executing cells.
How is the final score calculated?
The report assigns up to five points each for narrative, code, technical accuracy, and actionability.
Does it detect hardcoded secrets?
It runs detect-secrets when available and includes basic pattern matching as a fallback.
Can it verify every technical claim?
No. The audit identifies likely issues, but maintainers must verify behavior, APIs, and claims.
What local tools are required?
The validator expects Python and uv. It uses nbconvert for Markdown conversion and detect-secrets for credential scanning.
Developer Details
Author
anthropicsLicense
MIT
Skillstore revision
r1
Version notice
The author did not declare a version.
Ref
d71c7417a35d5c2624161bd2fe8de8a41a362128
Maintenance freshness
7/18/2026
Usage
5 downloads · 213 views
File structure