Back to skill

Security audit

Post Summarizer

Security checks for vulnerabilities and agentic risk

Overview

This skill is a news and article summarizer that uses web search and web fetching in a disclosed, purpose-aligned way.

Install this if you want a Chinese-language news/article summarizer that can browse public news sites. Be aware that broad prompts like asking what happened today may cause it to fetch several external sources, and outputs are instructed to be in Simplified Chinese unless the skill is changed.

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
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill description includes broad natural-language triggers such as '帮我看XX新闻', '总结这篇', and '最新XX资讯', which can overlap with ordinary user requests and cause the skill to activate when the user did not explicitly intend to invoke it. Because the skill performs web access and content retrieval, accidental invocation can lead to unnecessary external fetching, context hijacking from unrelated prompts, or surprising behavior.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The multiple mode trigger definitions are loosely separated and use overlapping phrases like '帮我看XX', 'XX热点', '搜索XX新闻', and '对比这几篇', making it unclear which mode should run under mixed or partial user input. This ambiguity increases the chance of unintended tool use, incorrect source selection, or execution of broader web searches than the user expected.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

Quick-brief triggers like '快报', '最新十条', and especially '今天发生了什么' are common conversational phrases that may appear in many contexts beyond financial/news briefings. Since this mode fans out across multiple sites and aggregates several items, accidental activation can amplify unintended browsing and produce broad external-data retrieval from a casual prompt.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The instruction "所有输出用简体中文" forces a specific language for all responses, including content from English-language sources, without offering the user a language preference. This is a language-policy issue because it removes locale choice rather than making Chinese a default or optional behavior.

Content

No source excerpt is available for this finding.