ClawHub
Back to skill

Security audit

restaurant-food-delivery-research

Security checks across malware telemetry and agentic risk

Overview

The skill is mostly transparent, but its helper script can call unrelated Crawlora APIs beyond restaurant and food-delivery research.

Review this before installing if you want a strictly food-only skill. It requires a Crawlora API key and sends your lookup inputs, including locations or grocery searches, to Crawlora. The main concern is not hidden malware, but that the included helper can be reused for broader Crawlora endpoints outside the advertised restaurant and delivery use case.

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 (3)

Description-Behavior Mismatch

Medium
Confidence
90% confidence
Finding
The helper is documented and implemented as a generic Crawlora API client rather than a narrowly scoped restaurant/food-delivery tool. That broad capability allows callers to access unrelated endpoints, expanding the skill beyond its declared purpose and increasing the chance of unauthorized data access or policy bypass through arbitrary web research.

Context-Inappropriate Capability

Medium
Confidence
94% confidence
Finding
The script forwards a caller-controlled path directly into the request URL with no validation, enabling access to any Crawlora endpoint the API key can reach. In the context of a narrowly described food-delivery research skill, this creates a scope-expansion primitive that can be abused for unrelated web research or data retrieval.

Missing User Warnings

Medium
Confidence
82% confidence
Finding
The script sends user-provided query/body data together with an API credential to an external third-party service, but provides no user-facing disclosure, confirmation, or guardrails around what data is being transmitted. This is risky because users may supply sensitive location, shopping, or personal preference data without realizing it is being sent off-platform to Crawlora.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

Detected: suspicious.exposed_secret_literal

File appears to expose a hardcoded API secret or token.

Critical
Code
suspicious.exposed_secret_literal
Location
SKILL.md:23