ClawHub

Sentiment Score

Security checks across malware telemetry and agentic risk

Overview

The skill appears to do sentiment scoring, but its server mode exposes an unauthenticated network endpoint that can use the installer’s OpenAI key.

Install only if you are comfortable sending submitted text to OpenAI and paying for usage on the configured API key. Prefer CLI mode or change the MCP server to bind to localhost and put it behind authentication before running it on any reachable host. Verify the Expanso tooling and remote deployment URL before using the deploy command.

SkillSpector

By NVIDIA
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 (3)

Missing User Warnings

High
Confidence
96% confidence
Finding
This pipeline forwards raw stdin content directly to the OpenAI API, which can expose sensitive user-provided data to an external service without any disclosure, consent flow, or input restrictions. Because stdin may contain arbitrary pasted text, users could unintentionally send secrets, personal data, or proprietary content off-system.

Vague Triggers

Medium
Confidence
89% confidence
Finding
The skill exposes a public HTTP POST endpoint at /score and binds the service to 0.0.0.0 without any visible authentication, authorization, origin restriction, or other invocation controls. That makes the endpoint reachable by any network client that can access the host, enabling unauthorized use, resource abuse, and cost amplification against the downstream LLM API.

Missing User Warnings

Medium
Confidence
93% confidence
Finding
The pipeline forwards user-supplied text directly to OpenAI using the configured API key, but there is no manifest warning, consent notice, or disclosure that submitted content leaves the local system and is processed by a third-party service. This creates a data exposure risk because users may send sensitive or regulated text under the assumption that processing is local or internal.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal