My Agent Browser

Security checks across malware telemetry and agentic risk

Overview

This appears to be a coherent browser automation/recording skill, but its recordings, traces, and saved sessions can contain sensitive login data.

Install only if you are comfortable letting the skill record browser workflows and reuse logged-in browser sessions. Treat generated recordings, traces, screenshots, and session-state files like secrets: store them privately, do not commit or share them, avoid recording sensitive accounts unless necessary, and delete artifacts when finished.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill explicitly documents saving and loading authenticated browser session state without warning that the state file may contain cookies, tokens, or other reusable authentication artifacts. In an agent context, this can lead to accidental persistence, reuse, or exfiltration of sensitive credentials across tasks or users.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The recording and tracing features capture page contents and explicitly preserve cookies/storage from the active session, yet the skill provides no privacy or data-handling warning. This creates a realistic risk of recording sensitive data, authenticated content, or tokens into artifacts that may later be shared or retained insecurely.

VirusTotal

61/61 vendors flagged this skill as clean.

View on VirusTotal