ClawHub
Back to skill

Security audit

Compliance Posture Intake

Security checks across malware telemetry and agentic risk

Overview

This appears to be a legitimate HIPAA posture intake, but it asks for sensitive compliance materials and creates a report without enough privacy, storage, or source-scope warnings.

Review before installing. Use redacted or summarized documents where possible, avoid patient identifiers unless absolutely necessary, verify any state-law output with qualified counsel, and decide where the generated .docx will be stored and deleted. Be aware that report generation may involve a separate docx skill or available Rote tooling even though that dependency is not clearly declared in metadata.

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 (4)

Context-Inappropriate Capability

Medium
Confidence
88% confidence
Finding
The skill directs the agent to perform open-ended web searches about state privacy and breach laws using broad queries derived from user inputs. That expands the skill from a bounded HIPAA intake into external legal research, which can introduce untrusted content, hallucinated legal conclusions, and scope creep without adequate guardrails or source validation. In a compliance context, users may over-rely on these outputs for legal obligations.

Description-Behavior Mismatch

Medium
Confidence
81% confidence
Finding
The skill promises broad compatibility but requires producing a .docx via an undeclared external 'docx skill', creating a hidden dependency not covered by the declared tool/interface model. In agent environments, undeclared capability requirements can cause fallback behavior, unexpected tool chaining, or silent handoff of sensitive compliance data to another component the user did not authorize or expect.

Missing User Warnings

Medium
Confidence
92% confidence
Finding
The skill instructs creation of a polished .docx report containing compliance posture, possible incident-response weaknesses, vendor/BAA gaps, and potentially health-data-related details, but it gives no warning about file creation, persistence, or handling of sensitive contents. In many agent contexts, writing such a file may store sensitive operational or regulated information on disk or in synced workspaces where retention and access controls are unclear.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The skill asks users to upload policies, BAAs, risk assessments, training records, and licenses for analysis without a clear privacy notice or data-minimization warning. Those documents can contain PHI, employee data, contract terms, security controls, and breach-response details; soliciting them without explicit caution increases the chance of unnecessary disclosure into the agent environment.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.