ClawHub
Back to skill

Security audit

AI Tech Lead

Security checks for vulnerabilities and agentic risk

Overview

This is a plain Markdown workflow skill for coding agents that enforces staged research, design, planning, and implementation with human review gates, with no hidden code or install-time behavior found.

Install this only if you want a strict architecture-first coding workflow. Before using it, decide whether you want the agent to ask separately before creating commits or spawning multiple sub-agents, since the skill treats those as part of its normal process.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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 (2)

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
*   **C4 Models** (Context, Containers, Components, Code).
*   **Data Flow** and **Sequence Diagrams** .
*   **ADR** (Architecture Decision Records) and testing strategies.
*   🛑 *Hard Stop: The process pauses here for mandatory human pair-review. The agent cannot proceed without approval*.

### 3. Planning 📝
The agent creates a detailed, step-by-step implementation plan. The plan is broken down into isolated phases (e.g., domain models, interfaces, adapters) with specific files targeted for creation or modification.
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.

Natural-Language Policy Violations

Low
Confidence
89% confidence
Finding
The README explicitly says the skill 'forces' the AI into a strict workflow, which reduces user choice and can override normal agent behavior without clear opt-in. In a skill for coding agents, coercive workflow control is risky because it can suppress user intent, constrain safer alternatives, or create prompt-level dominance over normal safety or usability expectations.

Static analysis

No suspicious patterns detected.