ClawHub
Back to skill

Security audit

Product & Innovation Playbook

Security checks for vulnerabilities and agentic risk

Overview

This is a product strategy playbook with broad but disclosed advisory instructions and no evidence of hidden execution, data access, or persistence.

Install this as an advisory product-management framework, not as an automation tool. Users who want narrow help in only one area, such as QA or competitive analysis, should invoke it deliberately and provide product stage, target segment, metrics, team capacity, and constraints so the broad playbook does not over-shape unrelated work.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (2)

Anti-Refusal Statement

High
Category
Anti-Refusal
Content
## 9) Output Format When Using This Skill

Always respond with:

1. Recommended approach
2. 30/60/90 execution plan
Confidence
84% confidence
Finding
The phrase 'Always respond with' is an anti-refusal / rigid formatting directive that attempts to constrain downstream agent behavior regardless of higher-priority context, safety requirements, or task suitability. While not overtly malicious here, such unconditional instructions can reduce the agent's flexibility to ask clarifying questions, decline inappropriate requests, or use a safer response structure when needed.

Vague Triggers

Medium
Confidence
95% confidence
Finding
The skill description defines very broad, overlapping trigger conditions spanning many common product and strategy tasks, which can cause the agent to invoke this skill in situations where a narrower or safer skill should apply. Overbroad activation increases the chance of incorrect delegation, instruction collisions, and unintended influence over unrelated workflows, especially because the playbook prescribes strong process defaults across discovery, QA, strategy, and R&D.

Static analysis

No suspicious patterns detected.