ClawHub
Back to skill

Security audit

Sync Notes

Security checks across malware telemetry and agentic risk

Overview

This is a real notes-sync skill, but it deserves review because it can change or delete local and cloud notes while storing cloud credentials in local skill files.

Install only if you are comfortable giving this skill access to your notes folder and the configured Cloudflare R2 bucket. Use a narrowly scoped R2 token, verify generated config files are not committed or backed up, run status and dry runs during initial setup, and remember that routine `/sync-notes` can modify or delete data on both local and remote sides.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill explicitly orchestrates shell execution (`bash`, `rclone`) but does not declare permissions or otherwise constrain that capability in metadata. In an agent environment, hidden or undeclared shell access increases risk because the model can trigger filesystem and network-affecting operations without an explicit trust boundary, especially for sync commands that read, write, and overwrite local data.

Credential Access

High
Category
Privilege Escalation
Content
├── config/
│   ├── rclone.conf           # active rclone config (created by setup, gitignored)
│   ├── rclone.conf.example   # template
│   ├── .env                  # non-rclone settings (created by setup, gitignored)
│   └── .env.example          # template
├── scripts/
│   ├── sync.sh               # main entry point (dispatcher)
Confidence
83% confidence
Finding
.env

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.