ClawHub
Back to skill

Security audit

antigravity-cli

Security checks across malware telemetry and agentic risk

Overview

This skill is documentation for using Antigravity CLI and its sensitive commands are disclosed and mostly guarded with safety guidance.

Before installing, verify the Antigravity CLI install source and avoid permission-bypass or always-proceed modes unless you are in a trusted, narrow workspace. Prefer sandboxing, review prompts, and the TUI permission controls for normal use.

SkillSpector

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

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The installation section recommends piping remote scripts directly into a shell (`curl | bash`, `irm | iex`) without any warning or integrity verification guidance. Even if the domain is intended to be official, this pattern trains users to execute network-fetched code blindly and increases the risk of compromise from DNS hijack, MITM, hosting compromise, or copycat documentation.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The flag `--dangerously-skip-permissions` is documented as auto-approving all tool requests without any warning, constraints, or usage guidance. In an agent CLI context, removing approval gates can allow file, shell, network, or plugin actions to proceed unchecked, materially increasing the blast radius of prompt injection, operator mistakes, or malicious tasks.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.