每日笔记深度解读

• Outcome: Review
• Latest audit: May 29, 2026 at 9:24 PM UTC
• Version:
  1.0.1

Overview

This notes skill is coherent, but it requires broad historical-note reading and full-text model analysis without clear opt-in or limits.

Install only if you are comfortable with the agent reading today's notes and searching older notes in the configured folder. Set NOTES_DIR to a narrow notes-only directory, exclude sensitive journals or private archives, and ask the agent to confirm before historical analysis or to use snippets instead of full note text.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The skill explicitly instructs searching across all historical notes and performing cross-note analysis without requiring explicit user consent, preview of scope, or minimization. In a notes skill, historical notes are likely to contain sensitive personal data, so broad collection and model exposure materially increases privacy risk beyond the user's immediate request to summarize today's notes.

Ssd 3

Medium

• Confidence: 97% confidence
• Finding: The skill directs the agent to search all historical notes and provide the full text of both today's notes and matched historical notes to the model for relevance ranking. This is a clear over-collection pattern that can expose large volumes of private user content to the model without necessity or explicit informed consent, especially since only top related notes are ultimately needed.

VirusTotal

64/64 vendors flagged this skill as clean.