ClawHub
Back to skill

Security audit

AI Tooling Inventory

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed healthcare compliance inventory workflow that asks users to identify AI tools and related metadata without requesting patient records or performing system changes.

Before installing, understand that this skill is designed for compliance discovery and may surface sensitive metadata about AI use, OAuth grants, browser extensions, and vendor AI features. Use it with an appropriate organizational respondent, avoid providing patient records, and treat any breach or BAA adequacy questions as matters for qualified counsel.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
75% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 20)May include surrounding context.

md
**Breach-determination boundary, state this in your output.** Path 5 findings — free tools used with a work email, browser extensions, consumer-tier signups — routinely prompt the question of whether a breach has already occurred. This skill does not make that determination. If a finding raises the question, flag it explicitly and direct the organization to seek qualified legal or regulatory counsel before concluding whether notification obligations apply. Think critically when raising this issue and present the specific facts used to determine the need to raise this flag. 

## The Governing Constraint: No Input May Require a Ticket
**Design rule:** the default run completes using only inputs the buyer persona can obtain under their own authority, in the same session, without asking anyone's permission. Elevated-access artifacts are an *optional upgrade pass*, never a prerequisite.

## The Central Problem: Four of Five Entry Paths Leave No Transaction

Static analysis

No suspicious patterns detected.