Security audit

create fun game

Security checks for vulnerabilities and agentic risk

Overview

The skill is a game-upload guide, but it asks agents to send local files to an external service and likely publish them without a clear final consent or safety-check boundary.

Install only if you intend agents to publish generated game files to the listed external service. Before using it, confirm the exact destination URL, inspect the three files, avoid absolute or user-supplied paths, check for secrets or private assets, and understand whether the resulting game is public and how it can be removed or overwritten.

A.I.G

Vulnerability Patterns

Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
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

Findings (1)

T09 · Insecure Skill Coding Practices

Warning

Location: SKILL.md:10
Finding: External File Upload and Publication Without an Explicit User-Consent Boundary
Content: View full analysis
**Repository URL**: https://www.idlab.top > Each game requires creating a new folder in the repository root directory, containing the following 3 fixed files: > - `config.json` — Game configuration (title and other metadata) > - `index.html` — Main game code > - `preview.png` — Game screenshot preview ``` It demonstrates uploading a file from a privileged local workspace path without requiring path validation, content review, or immediate user confirmation: ```bash curl -X POST "https://www.idlab.top/xlabopenapi/github/aigames/config" \ -F "gameFolder=xiaotongcoolgame" \ -F "file=@/root/workspace/logs/config.json" ``` The complete process instructs the agent to upload all three files: ```bash curl -X POST "https://www.idlab.top/xlabopenapi/github/aigames/config" \ -F "gameFolder=pixel_snake" \ -F "file=@config.json" ``` ```bash curl -X POST "https://www.idlab.top/xlabopenapi/github/aigames/index" \ -F "gameFolder=pixel_snake" \ -F "file=@index.html" ``` ```bash curl -X POST "https://www.idlab.top/xlabopenapi/github/aigames/preview" \ -F "gameFolder=pixel_snake" \ -F "file=@preview.png" ``` It then directs the agent to advertise a public-facing URL on a different domain: ```markdown ### Step 4: If successful, inform the user that the upload is complete Provide the experience URL as https://thenext.games/game/ + gameFolder. For example, in the case above: https://thenext.games/game/pixel_snake ``` ### Technical Analysis The skill's stated purpose is to guide the creation of mobile-compatible H5 games, but its prescribed workflow also performs an externally visible network side effect: ...[truncated 3138 chars]
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
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
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (3)

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 94% confidence
Finding: The skill instructs users to upload local files to a third-party service (idlab.top) using direct curl examples, but provides no warning that this transmits local content outside the user's environment. In an agent setting, this can normalize exfiltration of workspace files to an external domain, especially when file paths like /root/workspace/logs/config.json are shown as acceptable inputs.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 89% confidence
Finding: The instruction 'use English' for gameFolder imposes a language constraint in natural language. The document does not explain why this locale restriction is required by policy or system limitations, nor does it present it as a user choice.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The phrase 'Use English letters' is a natural-language locale restriction. As written, it appears to mandate English rather than narrowly specifying an implementation-safe character set with a documented technical rationale.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.