Scrapling Install

• Outcome: Review
• Latest audit: May 28, 2026 at 11:11 AM UTC
• Version:
  1.0.0

Overview

This is a legitimate Scrapling web-scraping aid, but it includes powerful anti-bot, proxy, credential, and destructive reset examples that need review before installation.

Install only if you intentionally need advanced Scrapling scraping and can keep its use limited to sites you own or are explicitly authorized to access. Do not hardcode proxy, API, or account credentials; review any stealth or Cloudflare-solving use carefully; and treat checkpoint deletion snippets as destructive commands.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (5)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 84% confidence
• Finding: The skill exposes clear network-capable behavior and instructions for fetching arbitrary URLs, but no explicit permissions boundary is declared. In an agent ecosystem, undeclared network capability weakens user awareness and policy enforcement, increasing the chance that the skill is used to access external systems unexpectedly.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 95% confidence
• Finding: The description presents the skill as guidance-only, but the content includes direct executable commands, local CLI usage, Python code that performs scraping, stealth fetching, proxy rotation, and Cloudflare-solving behavior. This mismatch is dangerous because it can mislead reviewers and users about the skill's actual operational capability, reducing scrutiny around high-risk scraping and anti-bot evasion functionality.

Missing User Warnings

Medium

• Confidence: 89% confidence
• Finding: The login-session example instructs users to place usernames and passwords into scraping automation without an explicit warning about secret handling, storage, redaction, or safe input methods. In agent-assisted environments, such examples can normalize insecure credential use and lead to secrets being hardcoded, logged, or exposed through transcripts and tooling.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The document explicitly promotes stealth fetching and Cloudflare bypass capabilities without any guidance on authorization, terms-of-service, or legal and ethical boundaries. In a scraping-focused skill, this materially lowers the barrier to misuse for evading anti-bot protections and accessing sites that are attempting to restrict automated access.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The recipe shows `shutil.rmtree("./crawl_checkpoint")` as a reset step without any explicit warning that it irreversibly deletes local data. In an agent skill context, destructive filesystem operations are more dangerous because users may copy them into automation or run them with modified paths, causing unintended data loss.

VirusTotal

66/66 vendors flagged this skill as clean.