ClawHub

ChainaAware Behavioral Prediction

Security checks across malware telemetry and agentic risk

Overview

The skill is a mostly disclosed ChainAware Web3 risk-analysis integration, but its API-key setup guidance and high-impact wallet profiling use cases need review before installation.

Install only if you intend to use ChainAware's external prediction service. Use a restricted or dedicated API key, prefer header/env-based integrations, and avoid browser URL setups where the key appears in history or network tools. Treat wallet scores as advisory signals, especially for lending, onboarding, AML, airdrops, or other decisions that affect users.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Intent-Code Divergence

Medium
Confidence
96% confidence
Finding
The skill makes inconsistent API-key handling claims: it says keys are never included in URLs or written to disk, yet other sections instruct users to put the key in query strings for Claude Web/ChatGPT and in config files such as Cursor JSON. This can cause operators to deploy the integration under false assumptions, increasing the chance of credential exposure via browser history, network tooling, logs, screenshots, backups, or checked-in config files.

Vague Triggers

Medium
Confidence
87% confidence
Finding
The manifest description contains a very broad catch-all trigger covering nearly any request about wallets, contracts, tokens, AI agents, trustworthiness, risk, behavior, or integration. Overbroad activation can cause the skill to run in contexts the user did not clearly intend, resulting in unnecessary transmission of wallet/contract identifiers to an external service and expanding the blast radius of any downstream mistakes.

Natural-Language Policy Violations

Medium
Confidence
85% confidence
Finding
The skill promotes profiling, segmentation, onboarding decisions, lending terms, bot filtering, and other forms of differential treatment based on predictive wallet analytics, without requiring meaningful opt-in beyond address-sharing consent. In context, this can enable opaque automated decision-making that affects access, pricing, or treatment of users, creating privacy, fairness, and compliance risk.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal