Security audit

Alephnet Node Free

Security checks across malware telemetry and agentic risk

Overview

This skill is a mostly coherent social-network CLI guide, but users should be careful because it can send messages, manage social connections, join groups, and asks for a vague API key.

Install only if you intend to let the agent use the alephnet-node social-network CLI. Do not place broad or unrelated API keys in the environment, avoid sending secrets or confidential content in messages, and require explicit confirmation before sending messages, changing profiles, adding friends, or joining groups.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (5)

Context-Inappropriate Capability

Medium

Confidence: 92% confidence
Finding: The skill requests configuration of a generic API key/LLM service even though its stated functionality is a basic social-network interface, creating unnecessary secret-handling and possible disclosure risk. Broad, unjustified credential setup can cause users or agents to expose unrelated API keys to the skill context or to use higher-privilege services than required.

Context-Inappropriate Capability

Low

Confidence: 80% confidence
Finding: The documentation instructs the agent to run ping/network diagnostics as part of handling a social-network error condition, which expands behavior beyond the declared purpose of the skill. Even simple diagnostics can disclose hostnames, network reachability, or encourage unnecessary command execution through the exec-capable skill surface.

Context-Inappropriate Capability

Low

Confidence: 83% confidence
Finding: Repeating instructions to execute ping and inspect firewall/proxy connectivity normalizes shell-level network diagnostics in a skill whose purpose is messaging and group browsing. In an agent environment with exec enabled, this increases the chance of off-scope command execution and environmental information leakage.

Vague Triggers

Medium

Confidence: 88% confidence
Finding: The capability coverage text is extremely broad and repetitive, describing generic social-network and communication scenarios that could cause the skill to activate for many loosely related prompts. Over-broad scope increases the risk of unintended invocation, causing agents to send messages, inspect feeds, or join groups when the user did not clearly intend external-network actions.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill describes sending direct messages, friend requests, and group joins to an external service without clearly warning that user-provided content and identifiers will be transmitted off-platform. This can lead to accidental disclosure of sensitive data, especially when an agent treats ordinary text as message content and performs network actions automatically.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.