Security audit

cctv-news-fetcher

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to be a CCTV news fetcher, but it asks for command execution and describes broad scraping, file handling, and automation beyond that narrow purpose.

Review this skill carefully before installing. It may be appropriate only if you intentionally want a broad agent helper with command execution and generic scraping, not just a CCTV news-by-date fetcher. Require clear command allowlists, source limits, credential handling, and user approval boundaries before granting exec or broad file access.

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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)

Description-Behavior Mismatch

Medium

Confidence: 93% confidence
Finding: The skill is presented as a narrowly scoped CCTV-news-by-date fetcher, but the documentation broadens it into generic information collection, workflow automation, and content processing. This scope drift is dangerous because users and agents may grant permissions or invoke behavior beyond the declared purpose, increasing the chance of misuse or unsafe execution paths.

Description-Behavior Mismatch

Medium

Confidence: 95% confidence
Finding: The documented scenarios include arbitrary URL scraping and generic data parsing, which materially exceed a date-based CCTV broadcast fetcher. Expanding from a fixed-source retriever to open-ended web/content fetching increases exposure to untrusted input, policy bypass, and unintended data collection.

Description-Behavior Mismatch

High

Confidence: 98% confidence
Finding: The feature list claims file handling, external API access, and system command execution despite the skill's stated role being simple news retrieval and parsing. Such undocumented capability expansion can cause over-privileged deployment and enables dangerous operations that attackers could repurpose through prompt or task injection.

Context-Inappropriate Capability

High

Confidence: 99% confidence
Finding: System command execution is not justified for a CCTV news fetcher and materially raises the attack surface. In an agent context, any exec-enabled skill handling untrusted content can be coerced into running local commands, accessing secrets, or pivoting into the host environment.

Intent-Code Divergence

Medium

Confidence: 91% confidence
Finding: The document first presents a narrow single-source fetcher, then later advertises multi-source aggregation, semantic search, scheduling, export, and execution-related functionality. These inconsistencies undermine trust boundaries and make it difficult for reviewers or users to understand what privileges the skill actually needs or may attempt to use.

Static analysis

No suspicious patterns detected.