Security audit

TencentCloud OCR

Security checks across malware telemetry and agentic risk

Overview

This is a coherent Tencent Cloud OCR skill, but users should understand that images, PDFs, URLs, and resume contents may be sent to Tencent Cloud for processing.

Install this only if you are comfortable sending selected images, PDFs, image URLs, and resume contents to Tencent Cloud for OCR. Use a dedicated least-privilege Tencent Cloud API key, avoid committing or printing credentials, monitor quota and billing, and require confirmation before processing IDs, legal files, resumes, screenshots, or other sensitive documents.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (6)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 80% confidence
Finding: The skill uses environment-based cloud credentials but does not declare permissions or trust boundaries, which creates an opaque capability surface for secret access and external API use. In an agent ecosystem, undeclared access to env-backed secrets can lead to overbroad execution and make review, sandboxing, and least-privilege enforcement harder.

Vague Triggers

Medium

Confidence: 87% confidence
Finding: The auto-trigger logic includes broad keywords such as '识别', '文字', '提取', and '读取', which can overlap with normal conversation and cause unintended invocation. That can result in user content, attachments, or URLs being sent to Tencent OCR without clear intent, creating privacy and data-minimization risks.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The phrase 'any scenario involving general OCR text recognition' is overly vague and effectively unbounded, encouraging the agent to invoke the skill without clear policy limits. In context, this is more dangerous because the skill handles potentially sensitive documents like legal files and resumes, so mistaken routing can expose private data to an external provider.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The skill is designed to automatically process image URLs, uploaded images, PDFs, and resumes via a third-party cloud OCR service, but it does not clearly warn that user-provided content will be transmitted externally. Because the stated use cases include sensitive materials such as legal dossiers and resumes, lack of disclosure and consent materially increases privacy and compliance risk.

Missing User Warnings

Low

Confidence: 76% confidence
Finding: The documentation instructs users to store cloud API secrets in environment variables but does not clearly warn that these are sensitive credentials that must not be exposed, logged, committed, or echoed. While common practice, omission of handling guidance can lead to accidental credential leakage during setup, debugging, or repository sharing.

Missing User Warnings

Medium

Confidence: 91% confidence
Finding: The script transmits user-provided image content or URLs to Tencent Cloud for OCR processing, but it provides no explicit notice, consent flow, or privacy warning before sending potentially sensitive documents off-host. In an agent skill context, this is meaningful because users may submit IDs, legal files, resumes, or other confidential content without realizing it will be sent to a third-party cloud service.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.