ClawHub
Back to skill

Security audit

scrape-creator-profile

Security checks across malware telemetry and agentic risk

Overview

This is a coherent public profile scraping skill, but it is broad enough to trigger live scraping, browser-session use, third-party Apify processing, and saved exports without strong per-action controls.

Review before installing. Use it only for authorized public-profile lookups, avoid broad lead-list or monitoring workflows unless clearly appropriate, and require explicit confirmation before using logged-in browser sessions, Apify fallback, or saving scraped data to disk.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Description-Behavior Mismatch

Medium
Confidence
97% confidence
Finding
The script persists scraped profile JSON to local disk, which exceeds a scrape/extract-only capability and creates a data-retention surface that may not be disclosed by the skill manifest. Even though the data is supplied explicitly via CLI and the write is local, storing scraped data can introduce privacy, compliance, and unintended collection risks, especially if profiles include personal or sensitive metadata.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The trigger guidance is broad enough that the skill may activate for generic research or analysis requests, causing unnecessary scraping, browser use, or collection of profile data when the user did not explicitly ask for that behavior. Over-broad invocation increases the chance of unintended data collection and tool use, especially because the skill can access browser, web fetch, bash, and file write capabilities.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill instructs the agent to save scraped profile data to local disk without requiring a clear notice or confirmation about persistence. Silent local storage can create privacy, retention, and data-handling risks, particularly when aggregating multiple creator profiles or exporting comparison files that may persist beyond the immediate session.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The file documents a fallback workflow that sends profile-related input data to a third-party service (Apify) and explicitly requires an API token, but it provides no warning that user-supplied targets, URLs, or identifiers will be transmitted off-platform. In a scraping skill, this is material because users may expect the agent to fetch data directly; silent third-party routing increases privacy, compliance, and consent risk, especially if personal website URLs or profile identifiers are sensitive.

External Transmission

Medium
Category
Data Exfiltration
Content
## Base API call pattern

```bash
curl -s -X POST \
  "https://api.apify.com/v2/acts/{ACTOR_ID}/run-sync-get-dataset-items?token=$APIFY_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{INPUT_JSON}'
Confidence
91% confidence
Finding
This example shows a direct POST request to Apify that transmits input JSON and an API token to an external service. While external calls are expected in a scraping reference, the issue is that the skill normalizes third-party transmission without any consent, disclosure, or safeguards around what user-provided data may be included in INPUT_JSON.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.