ClawHub
Back to skill

Security audit

Expense Categorization

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed expense-categorization workflow, but users should keep human oversight around reimbursement approvals and QuickBooks posting.

Install only if you are comfortable letting the agent classify sensitive expense documents. Configure the approval thresholds to your company's policy, require audit logs or sampling for auto-approved reimbursements, and do not allow QBO posting or reimbursement payment without your intended human review step.

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 (2)

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
4. Export mapped transactions to QBO

**Expense report approval routing:**
- Auto-approve: policy-compliant, under $250, no flags
- Manager approval: $250–$2,500 or single flag
- Finance review: >$2,500, multiple flags, or restricted category
Confidence
93% confidence
Finding
The skill explicitly recommends auto-approving expenses under a threshold based on policy checks and absence of flags, which is autonomous financial decision-making. In an accounting and reimbursement context, this can be abused by submitting crafted low-value fraudulent or miscategorized expenses that evade detection, leading to improper approvals without human validation.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
| Amount | Approver | SLA |
|--------|----------|-----|
| < $250 | Auto-approve (policy-compliant) | Immediate |
| $250–$999 | Direct manager | 2 business days |
| $1,000–$4,999 | Department head | 3 business days |
| ≥ $5,000 | Finance + CEO | 5 business days |
Confidence
85% 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.

Static analysis

No suspicious patterns detected.