cursor-agent

Security checks across malware telemetry and agentic risk

Overview

Review before installing: this Cursor CLI guide is mostly coherent, but it encourages automation that can trust workspaces and apply code changes without enough user control.

Install only if you intentionally want Cursor CLI automation in trusted repositories. Review project rules and MCP config first, avoid --force unless changes are version-controlled and reviewed, do not automate workspace trust, and understand why SKILLBOSS_API_KEY is needed before providing it.

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
88% confidence
Finding
The documentation promotes `--force` auto-apply mode for code changes without a clear warning that it can modify files automatically and bypass user review. In an agent/automation context, this increases the risk of unintended, unsafe, or destructive edits being applied directly to a repository or build workspace.

Missing User Warnings

High
Confidence
97% confidence
Finding
The tmux automation guide instructs sending `a` to trust the workspace automatically, bypassing an important security decision without any validation or warning. Trusting an unreviewed workspace can enable execution of project-defined rules, configs, or agent behaviors from `.cursor/rules`, `AGENTS.md`, `CLAUDE.md`, and related files, making this especially dangerous in automated or untrusted repositories.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal