ClawHub
Back to skill

Security audit

Boxed Curl

Security checks for vulnerabilities and agentic risk

Overview

This looks like a legitimate sandboxed curl helper, but it should be reviewed because it downloads and runs an unpinned WASM component and can send user-provided headers and bodies to external hosts.

Review before installing. Use only with trusted destinations, avoid sending secrets unless you intend them to reach that host, and verify or pin the downloaded WASM component before allowing the agent to run it.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger phrases are broad and natural-language oriented, making it easy for ordinary user requests like 'fetch data' or 'send a secure HTTP request' to invoke a capability that can send network traffic to arbitrary external hosts. In this skill, that broad activation surface is more dangerous because the tool supports user-controlled URLs, headers, methods, and bodies, which can transmit sensitive data externally if invoked unintentionally.

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The skill markets itself as 'safe' and 'sandboxed' but does not clearly warn that it will transmit user-supplied URLs, headers, and bodies to external services. This can mislead users or higher-level agents into treating it as harmless local processing, increasing the risk of accidental exfiltration of credentials, personal data, or internal content.

Static analysis

No suspicious patterns detected.