ClawHub
Back to skill

Security audit

Quotes

Security checks for vulnerabilities and agentic risk

Overview

This skill is a local personal quote organizer with disclosed quote storage and delivery behavior, but users should configure any automatic delivery deliberately.

Install only if you are comfortable keeping personal quotes, conversation snippets, tags, and optional mood/context notes in ~/quotes/. Before enabling automatic delivery, choose explicit schedules, notification channels, and categories so private or emotionally sensitive quotes are not surfaced unexpectedly.

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
  • 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
Findings (4)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The description promises automatically surfacing meaningful words without specifying the exact criteria, consent model, or limits of that automation. This ambiguity can enable the agent to infer mood, context, or relevance from personal interactions and deliver content unexpectedly, creating privacy and consent issues.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The skill defines broad behaviors like surfacing relevant quotes and creating a workspace without clearly constraining when autonomous actions should occur. In practice, vague activation boundaries can cause the agent to act on inferred user state or ambient context rather than explicit requests, leading to unexpected data storage or unsolicited output.

Autonomous Decision Making

Medium
Category
Excessive Agency
Content
## Core Behavior
- User shares quote → save with context and tags
- User needs inspiration → surface relevant quote
- Automatically send quotes based on schedule/criteria
- Create `~/quotes/` as workspace

## File Structure
Confidence
88% confidence
Finding
The phrase 'Automatically send quotes based on schedule/criteria' authorizes autonomous decision-making without defining strict guardrails, approval requirements, or boundaries on what criteria may be used. In this context, the agent could independently decide when a user needs inspiration or which sensitive quotes to surface, making the behavior more privacy-impacting than a simple local notes tool.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill includes automatic notifications and quote delivery but does not warn the user about unsolicited behavior or the privacy implications of storing and using personal emotional or contextual data. Because the skill tracks favorites, moods, and personal conversations, unannounced notifications can expose sensitive interests or inferred emotional state to others who can view the device.

Static analysis

No suspicious patterns detected.