ClawHub
Back to skill

Security audit

Subagent Driven Development

Security checks for vulnerabilities and agentic risk

Overview

This markdown-only workflow skill coordinates coding subagents and review gates, with no hidden persistence, exfiltration, or automatic install behavior found.

Install this only if you want a development workflow that delegates planned coding work to subagents and performs multiple review passes. Review the plan before invoking it, because subagents may edit code, run tests, and create commits as part of the intended workflow.

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
  • 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
Findings (2)

Session Persistence

Medium
Category
Rogue Agent
Content
[Read plan file once: docs/plans/feature-plan.md]
[Extract all 5 tasks with full text and context]
[Create TodoWrite with all tasks]

Task 1: Hook installation script
Confidence
60% 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.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
- Let implementer self-report replace independent verification
- Start code quality review before spec compliance is ✅
- Move to next task while any review has open issues
- Accept implementer's test report without verification

**If subagent asks questions:**
- Answer clearly and completely
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.

Static analysis

No suspicious patterns detected.