Skills notebooklm-manager

📦

notebooklm-manager

Name: notebooklm-manager
Author: LeeJuOh

Content revision r2 High Risk 🌐 Network access⚙️ External commands📁 Filesystem access

Manage NotebookLM Research from Your AI Assistant

Managing multiple NotebookLM notebooks and complete research queries is repetitive. This skill organizes the registry and delegates browser queries with bounded coverage checks.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Stop for confirmation before installing.

Review the plan and obtain explicit user consent before changing files.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "notebooklm-manager" from https://skillstore.io/skills/leejuoh-notebooklm-manager.md and its manifest at https://skillstore.io/api/skills/leejuoh-notebooklm-manager/manifest. Verify the artifact. Stop and obtain explicit user consent before installing or changing files.

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.

Markdown detail GET /skills/leejuoh-notebooklm-manager.md Signed manifest GET /api/skills/leejuoh-notebooklm-manager/manifest Signed lockfile GET /api/skills/leejuoh-notebooklm-manager/lockfile

Test it

Using "notebooklm-manager". List my active NotebookLM notebooks.

Expected outcome:

Claude Code Documentation: identifier claude-docs; topics include CLI, plugins, MCP, and hooks.
API Research: identifier api-research; topics include authentication, rate limits, and integrations.

Using "notebooklm-manager". Search my notebooks for browser automation.

Expected outcome:

Best match: Chrome Automation Notes. Identifier: chrome-automation. Topics: Chrome MCP, selectors, extension setup, and troubleshooting.

Using "notebooklm-manager". Ask claude-docs about plugin hooks and installation scope.

Expected outcome:

Notebook: Claude Code Documentation. Answer: Plugin hooks run around tool calls, while installation scope determines whether data is shared or project-specific.

Security Audit

High Risk

v5 • 7/23/2026 Open versioned report

All 64 static alerts are false positives caused by documented NotebookLM URLs, Markdown code formatting, app-owned storage paths, and ordinary lookup behavior. Context review found browser-agent prompt injection exposure, an unvalidated data-directory pointer, and spoofable response markers.

Files scanned

407

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (3)

High

SKILL.md:47 SKILL.md:71-80 references/commands.md:18-32

Untrusted Values Cross Browser-Agent Boundary

The skill inserts a stored URL and user question into a free-form prompt for a browser-capable agent. Crafted values can add instructions and steer an authenticated browser outside the intended NotebookLM workflow.

Direct interpolation into a privileged agent prompt is explicit. The companion agent's unseen constraints may reduce exploitability, so confidence is high but not absolute.

Medium

SKILL.md:22-26 SKILL.md:31 SKILL.md:47

Unvalidated Data Directory Pointer

The skill trusts an absolute path from a fixed locator and applies it to later file operations. A tampered locator can redirect configuration and registry access outside plugin storage.

The instructions explicitly trust an absolute path for all later operations. No canonicalization, symlink check, or approved-root validation is specified.

Medium

SKILL.md:89-99

Reserved Marker Injection in Agent Responses

The workflow selects error and truncation branches by searching agent output for reserved strings. Notebook content containing those strings can spoof errors, suppress coverage, or mislead users.

The response table explicitly uses content markers for control flow. Exploitation depends on whether the agent preserves matching text from notebook answers.

Risk Factors

🌐 Network access (4)

references/commands.md:24 references/commands.md:50 references/schemas.md:15 references/schemas.md:39

⚙️ External commands (45)

📁 Filesystem access (12)

SKILL.md:22 SKILL.md:170 SKILL.md:174 SKILL.md:187 SKILL.md:192 SKILL.md:193 SKILL.md:22 SKILL.md:170 SKILL.md:174 SKILL.md:187 SKILL.md:192 SKILL.md:193

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.

Open versioned report

Copy report link

https://skillstore.io/skills/leejuoh-notebooklm-manager/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/leejuoh-notebooklm-manager/security.svg)](https://skillstore.io/skills/leejuoh-notebooklm-manager?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/leejuoh-notebooklm-manager?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/leejuoh-notebooklm-manager/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/leejuoh-notebooklm-manager.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

LeeJuOh. (2026). notebooklm-manager security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/leejuoh-notebooklm-manager/audits/5

BibTeX citation

@techreport{leejuoh-leejuoh-notebooklm-manager-2026,
  author = {LeeJuOh},
  title = {notebooklm-manager security audit report (audit version 5)},
  institution = {Skillstore},
  year = {2026},
  number = {5},
  url = {https://skillstore.io/skills/leejuoh-notebooklm-manager/audits/5},
  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: "notebooklm-manager security audit report (audit version 5)"
version: "unspecified"
type: report
authors:
  - name: "LeeJuOh"
date-released: "2026-07-23"
url: "https://skillstore.io/skills/leejuoh-notebooklm-manager/audits/5"
identifiers:
  - type: other
    value: "skillstore:leejuoh-notebooklm-manager:audit:5"
    description: "Skillstore immutable audit report identifier"

Skillstore Score

Why this score Evidence Confidence: Medium

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

Synthesize Research Topics

Query a research notebook across several topics and fill coverage gaps with focused follow-ups.

Find Technical Guidance

Search documentation notebooks and retrieve a covered answer for an implementation question.

Maintain a Notebook Catalog

Add, search, archive, restore, and remove notebooks while keeping searchable local metadata.

Try These Prompts

List Active Notebooks

List my active NotebookLM notebooks with each notebook name, identifier, and topics.

Search the Library

Search my NotebookLM library for {topic}, then show the best matching notebook identifiers and descriptions.

Add and Query a Notebook

Add {notebook_url} with Smart Add, then ask it: {question}. Keep the existing chat history.

Run Coverage Follow-Ups

Query {notebook_id} about {topics}. Check every topic, run at most {followup_count} focused follow-ups, then synthesize the results.

Best Practices

Use specific notebook identifiers and focused questions for more reliable answers.
Review condensed questions before relying on responses to large inputs.
Keep follow-up limits small and confirm before extending coverage analysis.

Avoid

Do not submit unrelated websites as NotebookLM notebook URLs.
Do not place secrets or privileged browser instructions inside notebook questions.
Do not clear chat history unless the user explicitly requests a fresh context.