Second Brain Builder

• Outcome: Pass
• Latest audit: Jun 10, 2026 at 9:44 AM UTC
• Version:
  0.1.0

Overview

This skill is a disclosed second-brain research and note-writing helper with bounded, annotated local edits and no hidden execution or exfiltration behavior found.

Install this only if you want an agent to search your local second-brain notes and add AI 注 sections to them. Set SECOND_BRAIN_WIKI_ROOT or SECOND_BRAIN_ROOT so it targets the intended vault, and ask for a preview or diff before writeback if you want tighter control over changes or output language.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (2)

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The trigger language is broad enough that ordinary user requests such as 'research', 'update', or 'enrich' a topic could activate the skill unintentionally. Because this skill can write back to local second-brain files, over-broad invocation increases the chance of unauthorized or surprising file modifications from ambiguous prompts.

Natural-Language Policy Violations

Medium

• Confidence: 87% confidence
• Finding: Defaulting to Chinese without user opt-in can cause output-language mismatch, reducing user understanding of what the skill is about to write or change. In a writeback-capable skill, that increases the risk that users approve or overlook edits they cannot easily review, though this is primarily a usability and consent issue rather than a direct security flaw.

VirusTotal

62/62 vendors flagged this skill as clean.