ClawHub
Back to skill

Security audit

People’s Court Dispute Intake

Security checks for vulnerabilities and agentic risk

Overview

This is a narrowly scoped dispute-intake skill that asks the agent to summarize evidence without filing cases, moving money, contacting others, or collecting secrets.

Before installing, understand that this skill is for organizing potentially sensitive dispute information. Use redacted summaries where possible, keep original evidence in your own secure system, and do not treat its readiness summary as legal advice or a guaranteed path to recovery.

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 (1)

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
## Protect sensitive information

- Never ask the user to paste private keys, seed phrases, API tokens, passwords, or authentication cookies.
- Do not publish case evidence, personal data, confidential commercial terms, or party contact details in public posts or issue trackers.
- Prefer redacted summaries for initial scoping. Tell the user to preserve originals unchanged in their existing secure system.
- Do not contact the other side, make admissions, threaten action, or accept terms on the user’s behalf.
Confidence
80% 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.