ClawHub
Back to skill

Security audit

Agent Orchestration Patterns

Security checks across malware telemetry and agentic risk

Overview

This is a documentation-only skill for coordinating AI models; its main risk is that examples may share work data with multiple model providers.

Before using this skill on real projects, confirm that the data being routed to each model provider or sub-agent is allowed to leave your environment. Redact secrets and private customer data, use approved providers, and keep API keys in environment variables or a secret manager rather than embedding them in files.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill explicitly describes sending task content, draft outputs, self-reviews, and context across multiple third-party models, but it does not warn users that sensitive prompts or project data may be disclosed to multiple external providers. In an orchestration skill, this omission is material because the whole pattern encourages broad replication of potentially confidential data, increasing privacy, compliance, and data-governance risk.

Missing User Warnings

Medium
Confidence
88% confidence
Finding
The ACPX configuration shows wiring external services with API keys and an orchestration server, but the skill does not pair this with a warning about external connectivity, credential handling, or the implications of routing data through those services. This can lead users to deploy configurations that expose sensitive workflows to third-party systems without adequate notice, review, or guardrails.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.