Security audit

Market News Analyst

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed market-news analysis workflow that uses web research and local reference notes, with no hidden execution, persistence, or credential handling found.

Install this if you want an English-language market-news analyst that searches the web and summarizes recent financial events. Review its outputs carefully because market analysis can be incomplete or wrong, and it should not be treated as personalized financial advice.

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

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Natural-Language Policy Violations

Medium

Confidence: 98% confidence
Finding: The skill hard-codes English-only thinking and output at the metadata level without user opt-in or fallback behavior. This can override user language preference and create instruction-priority conflicts that reduce transparency and usability, though it is not a direct code-execution or data-exfiltration issue.

Natural-Language Policy Violations

Medium

Confidence: 97% confidence
Finding: Repeated English-only mandates reinforce behavior that may cause the agent to ignore user language preference across collection and reporting steps. The repetition makes the constraint more sticky and more likely to dominate normal user-directed behavior, increasing the chance of misaligned outputs.

Natural-Language Policy Violations

Medium

Confidence: 96% confidence
Finding: The report generation section explicitly requires structured English output only, which can force final content into a language the user did not request. In a user-facing skill, this is a policy/compliance vulnerability because it may prevent accessible or expected communication and can cause the model to disregard user instructions.

Natural-Language Policy Violations

Medium

Confidence: 95% confidence
Finding: The quality standards and principles require consistent English throughout, extending the language restriction into normative guidance. This broadens the risk that any downstream execution of the skill will systematically ignore user locale or accessibility needs.

Natural-Language Policy Violations

Medium

Confidence: 99% confidence
Finding: The important notes reiterate mandatory English-only thinking and output files, making the restriction explicit and difficult to override. While not malicious, this creates a persistent instruction conflict that can degrade trust and correctness in multilingual contexts.

Static analysis

No suspicious patterns detected.