ClawHub
Back to skill

Security audit

Lotus Wisdom

Security checks for vulnerabilities and agentic risk

Overview

This is a Markdown-only guided reflection skill with no code execution, persistence, credential use, or data movement.

Install only if you want a deliberately slow, step-by-step reflection style for complex questions. Expect it to pause after each step and wait for you to continue; it does not appear to run code, access files, use credentials, or persist changes.

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 (3)

Vague Triggers

Medium
Confidence
87% confidence
Finding
Many tag descriptions use broad concepts like 'deep diving,' 'considering perspectives,' or 'making it practical,' which overlap substantially and do not clearly distinguish when one tag should be selected over another. This weakens control over agent behavior and makes prompt steering easier, especially in a skill that depends on precise sequential progression through steps.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The framework tells the agent to 'trust organic movement' and 'let the content guide the method' without concrete decision rules, stop conditions, or disallowed transitions. In an interactive step-by-step skill, this ambiguity can be exploited by adversarial or manipulative user input to steer the agent into inappropriate modes, bypass intended structure, or produce inconsistent behavior.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The activation guidance includes broad categories such as complex problems, creative blocks, strategic decisions, system design, and philosophical questions, which can match a large share of ordinary user conversations. That increases the chance the skill is invoked when not explicitly requested, potentially steering the assistant into a rigid stepwise interaction that disrupts normal behavior or overrides more appropriate handling.

Static analysis

No suspicious patterns detected.