ClawHub
Back to skill

Security audit

AANA Purchase Booking Guardrail Skill

Security checks for vulnerabilities and agentic risk

Overview

This is a protective instruction-only guardrail for purchases and bookings, with no code execution, persistence, or hidden install behavior found.

This skill is appropriate as a guardrail for agents that may shop, book, renew, cancel, donate, or transfer funds. Users should still ensure any configured external checker is trusted and that review payloads stay redacted as the skill instructs.

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 (4)

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
4. Check authorization: confirm the user explicitly approved this exact action and cost.
5. Check payment privacy: do not expose full payment numbers, bank details, credentials, or unnecessary account data.
6. Check reversibility: prefer cart, quote, draft, hold, or review screen before final submission.
7. Check scope: do not add extras, warranties, insurance, upsells, tips, donations, seats, bags, or subscriptions without approval.
8. Choose action: accept, revise, ask, defer, refuse, or route to human review.

## Required Pre-Flight Checks
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.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
4. Check authorization: confirm the user explicitly approved this exact action and cost.
5. Check payment privacy: do not expose full payment numbers, bank details, credentials, or unnecessary account data.
6. Check reversibility: prefer cart, quote, draft, hold, or review screen before final submission.
7. Check scope: do not add extras, warranties, insurance, upsells, tips, donations, seats, bags, or subscriptions without approval.
8. Choose action: accept, revise, ask, defer, refuse, or route to human review.

## Required Pre-Flight Checks
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.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
Prefer:

- compare options before choosing,
- prepare a cart without checking out,
- hold or reserve only when the terms are clear,
- ask the user to complete payment directly,
- use a review screen before final confirmation,
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.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
- If final commitment facts are incomplete, ask.
- If the draft includes unapproved add-ons, recurring terms, fees, or payment details, revise.
- If the action is high-value, non-refundable, legally binding, or for someone else, defer or require explicit approval.
- If the request would expose payment secrets, bypass consent, misrepresent terms, or spend money without approval, refuse and explain briefly.
- If a checker is unavailable or untrusted, use manual purchase and booking review.

## Output Pattern
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.