ClawHub
Back to skill

Security audit

design-thinking

Security checks for vulnerabilities and agentic risk

Overview

This design-thinking guide fits its stated purpose, with ordinary privacy precautions needed when researching real users.

Obtain participants' consent, anonymize research notes, and omit unnecessary personal or confidential business information from reports and reusable examples.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • 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)

The user-research workflow lacks participant-consent and privacy warnings.

Medium
Category
Output Handling
Confidence
88% confidence
Finding

The workflow instructs interviews and contextual observation of real users, then includes participants and behavioral findings in the report. It provides no warning to obtain informed consent or protect identifying and confidential information collected during research.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.