ClawHub
Back to skill

Security audit

Deep Thinking

Security checks for vulnerabilities and agentic risk

Overview

This is a reasoning-guidance skill with broad activation language but no hidden commands, data access, persistence, or external execution behavior.

Install this if you want the agent to spend more effort on careful analysis for complex tasks. Expect possible extra verbosity or token use because the activation scope is broad, but I found no evidence of hidden execution, data collection, credential use, or persistence.

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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill description says it 'automatically applies' to broad categories like multi-step problems, ambiguity, architectural decisions, debugging, and anything requiring careful analysis. That wide trigger surface can cause unintended invocation, overriding more specific skills or injecting verbose process guidance into tasks where it is unnecessary, which can alter agent behavior in unpredictable ways.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The 'When to Apply' section uses subjective criteria such as 'multiple valid approaches,' 'stakes are high,' and 'touches multiple systems or files' without clear boundaries or counterexamples. In an agent setting, ambiguous routing guidance increases accidental activation frequency and can lead to context pollution, excessive reasoning behavior, or misalignment with simpler task-specific skills.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
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.

Content

Scanner excerpt · reference.md (reported line 117)May include surrounding context.

md
- **Premature conclusions**: Settling on an answer before exploring alternatives
- **Overlooked alternatives**: Fixating on one approach because it's familiar
- **Logical inconsistencies**: Conclusions that contradict earlier findings
- **Unexamined assumptions**: Treating assumptions as facts without verification
- **Incomplete analysis**: Stopping at the first layer when deeper investigation is needed
- **Confirmation bias**: Only seeking evidence that supports your current theory

Static analysis

No suspicious patterns detected.