ClawHub
Back to skill

Security audit

Archive Free

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward local content archiver with disclosed local storage, though users should understand it may use agent/LLM processing and local file writes.

Install only if you are comfortable with the agent fetching or processing the URLs/text you provide, using an LLM/API for summaries and tags, and writing Markdown archive files under ~/archive/. Avoid using it for private documents or sensitive links unless your agent and provider configuration are appropriate.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (4)

Context-Inappropriate Capability

Medium
Confidence
92% confidence
Finding
The manifest grants `exec` even though the documented purpose is content archiving and search, and the file does not constrain what commands may be run. In an agent setting, generic command execution materially expands the attack surface because untrusted URLs, page content, or prompts could influence filesystem or shell actions beyond simple archiving.

Intent-Code Divergence

Low
Confidence
84% confidence
Finding
The dependency and API-key documentation understates the skill's real capabilities and trust boundaries: the workflow implies external content retrieval, local file creation, and command execution, but only an LLM API is disclosed. This mismatch can mislead users about what data leaves the system and what local actions the skill may perform, reducing informed consent and safe deployment.

Missing User Warnings

Low
Confidence
88% confidence
Finding
The skill description omits an explicit warning that using the skill will create local directories and save snapshots as files under `~/archive/`. In agent environments, undisclosed filesystem side effects can surprise users, lead to unwanted data persistence, and make it easier to hide unsafe or excessive local writes behind a seemingly simple archive feature.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill does not warn that archiving external URLs or pasted content may cause those inputs, and possibly extracted page text, to be transmitted to an LLM/API service. This is a meaningful privacy risk because users may archive sensitive links, private documents, or research materials without realizing they could be exposed to third-party processing.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.