ClawHub
Back to skill

Security audit

Legal Intake Pilot

Security checks for vulnerabilities and agentic risk

Overview

This legal-intake skill is coherent, but it should be reviewed because it collects, stores, and may share sensitive legal intake data without a clear upfront privacy and consent step.

Before installing, confirm that your firm adds an upfront privacy notice, obtains consent before saving or sending intake details to any CRM, limits what PHI/PII is collected, defines retention and access controls for the leads database, and localizes the legal disclaimer for users who do not operate in English.

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
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly instructs the agent to collect sensitive intake details, write them to a leads database, and push qualified leads to third-party CRM platforms, but it does not clearly warn end users at the point of interaction that their data will be stored and shared. In a legal-intake context, this can expose highly sensitive personal, medical, and adversarial-party information without informed consent, increasing privacy, confidentiality, and compliance risk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The rule hard-codes an English-only disclaimer requirement ('Not legal advice') for every first response and session close, without any provision for the user's language or accessibility needs. In a legal intake context, this can cause users with limited English proficiency to miss a critical limitation on the service, increasing misunderstanding risk and weakening the effectiveness of the safeguard.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

The instruction requires that every interaction MUST include one fixed English disclaimer: "I am an AI assistant, not an attorney. This is not legal advice." This can create a language-policy issue for users who do not operate in English because the skill does not offer localization or user opt-in for the forced language.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.