ClawHub

Add Literature

Security checks across malware telemetry and agentic risk

Overview

This is a disclosed literature-import skill that needs a platform API key and writes literature records, with some credential/TLS hygiene caveats to review.

Install only if you intend to let the agent use your platform API key to add literature records and optional open-access PDF artifacts. Prefer the public TLS endpoint or verify any internal self-signed certificate out of band before trusting it, and store the API key in your MCP client or secret manager rather than pasting real keys into shared scripts or shell history.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The document provides concrete instructions for using a bearer API key on every request, including example commands that may encourage operators to paste long-lived credentials into CLI history or code without any warning about storage, rotation, least privilege, or log exposure. In a skill meant to connect autonomous agents to a publishing platform, this increases the chance of credential leakage and subsequent unauthorized use of MCP tools.

Missing User Warnings

High
Confidence
98% confidence
Finding
The instruction to trust a self-signed certificate normalizes bypassing standard certificate validation without describing how to verify the certificate fingerprint or securely distribute a private CA. That can enable man-in-the-middle interception of bearer tokens and MCP traffic, especially because the same section requires Authorization headers on every request.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal