ClawHub
Back to skill

Security audit

Privacy Compliance Guide

Security checks for vulnerabilities and agentic risk

Overview

This is a static privacy-compliance guide with templates and checklists, not a skill that runs code or accesses data automatically.

Installers should treat this as practical compliance guidance, not legal advice. It may prompt discussion of customer data flows, vendors, revenue thresholds, and marketing practices, so only share details appropriate for the agent session and consult qualified counsel for binding privacy-law decisions.

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 (3)

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 15)May include surrounding context.

md
| Decision | Strong | Acceptable | Weak|
|---|---|---|---|
| **Privacy policy** | Custom-drafted, regulation-specific, regularly updated | Template-based, covers key regulations | Generic boilerplate or missing |
| **Cookie consent** | Granular opt-in banner with category controls | Basic opt-in/opt-out banner | No consent mechanism or implied consent |
| **Data inventory** | Complete map of all data collected, stored, processed, shared | Major data flows documented | No documentation of data practices |
| **Email compliance** | Double opt-in, easy unsubscribe, CAN-SPAM + GDPR compliant | Single opt-in with working unsubscribe | Purchased lists or no unsubscribe option |
| **Data retention** | Defined retention periods per data type with auto-deletion | General retention policy exists | No retention policy, data kept indefinitely |

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 227)May include surrounding context.

md
1. **Relying on "implied consent"** — Under GDPR, pre-checked boxes, continued browsing, or scroll-based consent are NOT valid. You need affirmative action (click "Accept") for non-essential cookies and marketing.

2. **Using purchased email lists** — This violates CAN-SPAM (if recipients haven't opted in) and GDPR (no consent basis). Delete purchased lists immediately and build organically.

3. **Firing tracking pixels before consent** — Many sites load Google Analytics and Facebook Pixel on page load, before the cookie banner is answered. This is a GDPR violation. Implement consent-gated loading.

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 · references/implementation-toolkit.md (reported line 78)May include surrounding context.

md
On marketplaces, the platform handles most customer-facing privacy (cookie consent, privacy policy for marketplace transactions). Your obligations:
- Maintain your own privacy policy for your seller profile/About page
- Comply with marketplace data use policies (don't use buyer data for unauthorized purposes)
- Don't export buyer data to external marketing lists without consent
- Handle any direct customer communications (support emails) in compliance with CAN-SPAM/GDPR
- Maintain DPAs with your own tools that process marketplace order data

Static analysis

No suspicious patterns detected.