Book Fetch

• Outcome: Review
• Latest audit: May 28, 2026 at 8:31 AM UTC
• Version:
  1.0.0

Overview

This skill mostly does what it advertises, but it can automatically download books from shadow-library sources and upload them to a MEGA account without a clear confirmation step.

Install only if you intentionally want an agent to contact Anna's Archive/Libgen, download files to /tmp/books, and upload them into MEGA. Use --dry-run for search-only use, use --pick -1 for ambiguous titles, verify which MEGA account mega-put or rclone is authenticated to, and confirm you have the right to download and store the requested books.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 96% confidence
• Finding: The skill explicitly performs outbound network access to Anna's Archive/libgen and cloud upload to MEGA, and it invokes command-line tooling such as rclone, yet no permissions or safety boundaries are declared. That mismatch is dangerous because an agent may execute file transfer and shell-capable actions without transparent user consent, review, or policy gating, increasing the risk of unauthorized downloads, data exfiltration, and abusive automation.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The trigger wording is broad enough to match ordinary requests like 'find a book' or 'download an ebook,' which can cause the skill to activate in many benign reading-related conversations. In this context, overbroad activation is especially risky because the skill's behavior is not passive search: it downloads potentially infringing material from shadow-library sources and uploads it to remote storage automatically.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The description does not clearly warn users that the skill will download files from third-party ebook piracy sources and then upload them to MEGA, creating both security and legal risk. Without an explicit warning, users may unknowingly authorize retrieval of untrusted files and transfer of content to cloud storage, which can lead to malware exposure, copyright infringement, and unintended persistence of downloaded material.

Missing User Warnings

Medium

• Confidence: 97% confidence
• Finding: The script automatically uploads downloaded content to MEGA without an explicit confirmation step or a clear warning that data is being transferred to a third-party cloud service. In an agent skill context, silent exfiltration to external storage is more dangerous because users may think the action is only downloading a file locally, not copying it to a remote account.

VirusTotal

62/62 vendors flagged this skill as clean.