ClawHub
Back to skill

Security audit

Multi-Agent Debate

Security checks for vulnerabilities and agentic risk

Overview

This skill is a markdown-only guide for using structured multi-agent debate and does not request system access or persistence.

Installers should expect higher model usage and latency when applying the debate pattern, because it simulates multiple independent answers and critiques. I found no hidden execution, persistence, credential handling, or broad local access in the inspected artifacts.

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
  • Rogue AgentSelf-Modification, Session Persistence
  • 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)

Session Persistence

Medium
Category
Rogue Agent
Confidence
75% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · references/examples.md (reported line 39)May include surrounding context.

md
- Self: Good coverage, but should emphasize Bell got the patent that mattered.

**Agent C Critiques:**
- Agent A: Too simplistic. "Invented" is loaded term.
- Agent B: Most accurate about complexity of invention.
- Self: I missed Meucci. Should acknowledge prior work.