Browser Automation

• Outcome: Review
• Latest audit: May 28, 2026 at 7:08 AM UTC
• Version:
  0.1.0

Overview

This appears to be a legitimate Chrome automation skill, but it gives AI clients broad access to a real logged-in browser without enough scoping or consent guidance.

Install only if you trust the upstream bridge and Chrome extension. Prefer a dedicated Chrome profile, avoid using sensitive logged-in accounts, pin and verify the npm package and extension release where possible, and require explicit confirmation before the agent reads history, captures network traffic, sends cookie-backed requests, submits forms, deletes bookmarks, posts content, or changes account data.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• 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

Findings (4)

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The top-level description is broad enough to trigger on many ordinary browsing-related requests without requiring explicit user confirmation for sensitive actions. In a skill that can access an existing Chrome profile, sessions, history, bookmarks, and network data, overbroad activation materially increases the chance the agent invokes powerful browser capabilities in situations where the user did not intend that level of access.

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The "When to Use" section lists many capabilities but provides no guardrails, exclusions, or approval requirements for privacy-sensitive actions. Because this skill operates against the user's real Chrome environment, missing trigger constraints can cause the agent to over-apply the skill and access data or perform actions beyond the user's expectations.

Missing User Warnings

High

• Confidence: 97% confidence
• Finding: The skill advertises use of existing Chrome login sessions and later documents access to browsing history, bookmarks, and network-capable features, but it does not prominently warn that these are privacy-sensitive and may expose authenticated content or personal data. In this context, the omission is especially dangerous because users may interpret ordinary browser help as low risk while the skill can act with the full authority of their current browser state.

Missing User Warnings

High

• Confidence: 98% confidence
• Finding: Documenting network capture and cookie-backed HTTP requests without an explicit warning understates the risk of exposing session tokens, authenticated traffic, CSRF-protected actions, or other sensitive request metadata. This is more dangerous here because the tool runs against the user's live browser context, so misuse could reveal or act on highly privileged data from already-authenticated services.

VirusTotal

62/62 vendors flagged this skill as clean.