ClawHub
Back to skill

Security audit

Deck Builder

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent PowerPoint deck generator, with disclosed but resource-heavy slide delegation that users should control.

Install only if you want an agent to create full PowerPoint decks, including many slide-generation subtasks. Before using it, specify the desired slide count and confirm whether automatic PPTX assembly and submission are acceptable for your workflow.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (4)

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest presents the skill as a generic 'deck builder' but the instructions authorize materially broader behavior: orchestrating multiple subordinate agents, generating many slide tasks, assembling a PPTX, and submitting the result. That mismatch reduces transparency and can cause users or supervising systems to invoke the skill without understanding its full execution and delegation scope.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill lacks clear activation criteria and scope boundaries, while containing broad instructions to generate 15+ slides, delegate to many subordinate agents, and assemble outputs. Without explicit triggers and limits, the skill is more likely to run in unintended contexts, over-consume resources, or perform actions beyond user expectations.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The instruction text says 'You are a senior PPT Architect (Deck Builder),' which hard-codes the assistant's role and presentation behavior without offering any user choice on language or locale preferences. While not explicit about a locale, the skill imposes a fixed mode of interaction and output assumptions without documenting any opt-in or region-specific need.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
76% confidence
Finding

Building a presentation deck reasonably justifies creating slide plans and assembling a PPTX, but the explicit instruction to call submit_result adds an outward completion/submission action not described in the manifest. That capability changes the skill from content creation into workflow completion on behalf of the caller.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.