Security audit

Code Detective

Security checks for vulnerabilities and agentic risk

Overview

This is a Russian-language debugging game skill with no executable code or persistence, though its broad trigger wording could confuse game play with real debugging help.

Install this if you want a Russian-language Code Detective game. If you often ask the assistant to debug real code in Russian, be explicit when you want actual debugging help rather than the game, because the trigger phrase "найди баг" is somewhat broad.

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

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
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
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (4)

Ae4

Medium

Category: analysis-evasion
Confidence: 80% confidence
Finding: Suspicious Unicode normalization or mixed-script content

Vague Triggers

Medium

Confidence: 94% confidence
Finding: The activation guidance is broad enough to match ordinary debugging requests such as "найди баг" or "потренироваться в отладке," which can cause the game skill to trigger when the user actually wants real debugging help. This creates an instruction-routing and user-intent integrity problem: the assistant may divert from the requested task, provide toy examples instead of analyzing the user's code, or mishandle sensitive code pasted for actual support.

Natural-Language Policy Violations

Medium

Confidence: 88% confidence
Finding: The skill metadata and content prescribe Russian-language interaction without any user-language negotiation or fallback. While not a classic exploit, this can impair safe and accurate operation by overriding user preferences, increasing misunderstanding risk, and reducing the user's ability to understand prompts, buttons, and guidance.

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The usage section repeats ambiguous invocation conditions without clear boundaries, reinforcing that the skill may activate on generic debugging language rather than explicit opt-in to a game. In context, this is more dangerous because the skill is user-invocable and interactive, so an accidental trigger can derail legitimate assistance and lead to confusing or inappropriate responses.

Static analysis

No suspicious patterns detected.