ClawHub
Back to skill

Security audit

Image To Text Pdf

Security checks for vulnerabilities and agentic risk

Overview

This skill appears to be a straightforward local tool for making searchable PDFs from images, with no hidden network, credential, persistence, or destructive behavior found.

Install only if you are comfortable running local PDF/OCR helper scripts on the documents you provide. Avoid using --check-text on sensitive documents in logged terminals, and review the debug PDF to confirm the embedded copyable text matches your intent before sharing the final PDF.

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
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The declared description is for an end-user conversion tool that takes a finished raster image and outputs one or more PDFs with searchable text overlays. The supplied code chunk is a narrower preprocessing utility: it transforms OCR word-box data into line-level layout JSON. While this could support a larger pipeline for creating searchable PDFs, by itself it does not perform the core declared actions of PDF generation, text-layer embedding, or inspection-PDF creation. The code’s actual inputs/outputs and primary purpose therefore differ materially from the declared description.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The docstring says the script converts OCR word boxes into layout JSON, suggesting a direct transformation from OCR input to layout output. However, source_lines and correct_with_source allow the output text to come from a separate source file instead of the OCR words themselves.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest describes converting a finished raster image into a PDF text layer, which implies preserving OCR-derived text placement. This script can instead replace recognized line text with best-matching lines from a separate --source-text file, making it a text correction/reconstruction tool rather than only an OCR-layout converter.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The --check-text option prints all extracted text from the generated PDF directly to standard output. Although the flag name hints at text checking, there is no explicit warning that potentially sensitive document contents will be echoed to the console, logs, or calling environment.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.