ClawHub
Back to skill

Security audit

agent-desktop-ffi

Security checks across malware telemetry and agentic risk

Overview

This is a developer reference skill for a desktop-automation FFI; it exposes powerful screen and UI-control capabilities, but they are disclosed and aligned with the stated purpose.

Install this only if you intend to let agents or applications inspect and control the desktop through the host process. Treat snapshots, screenshots, clipboard reads, logs, and last-error details as sensitive, and enable tracing only when you want session files written under ~/.agent-desktop.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Rogue AgentSelf-Modification, Session Persistence
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Missing User Warnings

Medium
Confidence
87% confidence
Finding
This markdown file includes worked examples that snapshot the focused window, print captured JSON, and execute UI actions like clicking and typing text. Those behaviors can affect user data, privacy, and system integrity, but the document does not include a clear warning or disclosure to users before demonstrating them.

Session Persistence

Medium
Category
Rogue Agent
Content
- **Structured session trace (no ABI change).** File-based JSONL tracing activates
  only when the session has a manifest with `trace: on` from `session start`
  (CLI) or equivalent on-disk setup. `ad_adapter_create_with_session` alone does
  **not** create trace files. When tracing is active, `command_context()`-backed
  commands append to one segment per OS process under
  `~/.agent-desktop/sessions/<id>/trace/<pid>-<procTs>.jsonl`. A long-lived host
  reuses the same segment filename for all calls in that process. For unstructured
Confidence
60% confidence
Finding
Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.