Ae1
High
- Category
- analysis-evasion
- Confidence
- 100% confidence
- Finding
Referenced artifact was not completely inspected
- Content
md - `references/international-search.md` - International search guide
Security audit
Multi Search Engine.Tmp
Security checks for vulnerabilities and agentic risk
Overview
This is a static search-helper skill, but it includes unsafe search examples for finding exposed passwords or deleted content and for disabling safety filters without warnings.
Review this before installing if your environment has rules about acceptable search behavior. The skill does not install code or request credentials, but its guide includes examples that could be misused to search for exposed secrets or unsafe content. If used, keep queries user-directed, avoid credential-hunting patterns, and prefer safe-search defaults.
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
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
- Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (7)
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 93% confidence
- Finding
The visible documentation sections and usage context are presented in Chinese only, with no indication that users may choose another language or that the skill is intentionally limited to a Chinese-speaking audience. This can violate a language/locale policy when the skill is exposed to a broader user base without opt-in.
- Content
Natural-Language Policy Violations
Medium
- Category
- Not specified by scanner
- Confidence
- 90% confidence
- Finding
SQP-3 applies to natural-language policy violations across all file types, including language/locale constraints imposed without user opt-in. This file presents all instructions exclusively in Chinese with no indication that the skill is China-specific or that alternate language support is unavailable by design.
- Content
Ssd 2
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The examples include dork-like queries such as searching for 'intext:password filetype:txt' and viewing cached/deleted content, which can facilitate discovery of exposed credentials or sensitive remnants on third-party systems. In a search-integration skill, this is more dangerous because it operationalizes reconnaissance patterns users could directly reuse through the agent.
- Content
Context-Inappropriate Capability
Medium
- Category
- Not specified by scanner
- Confidence
- 95% confidence
- Finding
The manifest describes a skill focused on integrating multiple search engines, advanced search operators, filters, privacy engines, and WolframAlpha queries. This section documents non-search utility features exposed via DuckDuckGo queries, which are not necessary to implement or explain search-engine integration and materially broaden the skill's practical capabilities.
- Content
Missing User Warnings
Medium
- Category
- Not specified by scanner
- Confidence
- 94% confidence
- Finding
The guide explicitly documents parameters that disable SafeSearch/family filtering without any warning about adult, unsafe, or workplace-inappropriate content. In an agent skill, this can lead to unreviewed retrieval of explicit or harmful content, especially if user input is incorporated into generated search URLs automatically.
- Content
Missing User Warnings
Low
- Category
- Not specified by scanner
- Confidence
- 76% confidence
- Finding
The privacy best-practices section advises '使用Tor浏览器 + DuckDuckGo onion服务' for sensitive searches, but does not disclose that Tor usage may violate workplace network policy, trigger monitoring, or have operational constraints. In markdown guidance, recommending privacy-impacting or network-sensitive behavior should be accompanied by a user warning.
- Content
Static analysis
No suspicious patterns detected.