ClawHub
Back to skill

Security audit

PPT from Template

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent presentation-template helper that uses local conversion and extraction tools without hidden network access, credential use, destructive behavior, or system persistence.

Installers should understand that this skill will scan the workspace template folder, save uploaded PPTX templates there for reuse, and run local document-conversion tools on presentation files. Use it with templates you trust and confirm the template selection when a generic PPT request might be interpreted as template-based generation.

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
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (4)

Tp4

High
Category
MCP Tool Poisoning
Confidence
98% confidence
Finding

The declared description focuses on style extraction from reference templates and generation of new presentation slides from scratch. The actual code does not analyze templates, extract style, generate slides, use PptxGenJS, handle PDFs as style references, or support template selection/placeholders/batched deck creation. Instead, it performs a different task: exporting an existing PPTX to PDF through a headless LibreOffice invocation. That is a materially different primary purpose, so this is a clear description-behavior mismatch.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding

The skill instructs the agent to read and write workspace files and execute shell commands, but it declares no tool scope or allowed-tools boundary. That creates an over-privileged execution model where the runtime may permit broader filesystem or shell access than the skill actually needs, increasing the chance of unintended file access or command abuse if inputs or downstream scripts are compromised.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger phrases are broad enough to match generic requests like '做个PPT' or 'make slides like this,' which can cause the skill to activate in situations where the user did not intend template-based file processing. In context, this matters because activation can lead to file scanning, uploaded document handling, and shell-driven conversion workflows, expanding exposure to risky operations unnecessarily.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/pptx_to_pdf.py (reported line 35)May include surrounding context.

python
cmd = [sys.executable] + cmd

        print(f"🔄 Converting: {os.path.basename(pptx_path)} → PDF")
        result = subprocess.run(cmd, capture_output=True, text=True, timeout=120)

        if result.returncode != 0:
            print(f"❌ Conversion failed:\n{result.stderr}", file=sys.stderr)

Static analysis

No suspicious patterns detected.