Security audit

Drama Script

Security checks for vulnerabilities and agentic risk

Overview

This is a Chinese short-drama writing workflow that creates and manages local project files, with no evidence of hidden code execution, credential access, exfiltration, or destructive behavior.

Install only if you want a Chinese-language short-drama workflow that writes structured project files in the current workspace. Use a dedicated workspace if you do not want it scanning for existing drama drafts or projects, and be aware that its character guidance includes a rigid binary gender-balance rule.

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
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 (9)

Natural-Language Policy Violations

Medium

Confidence: 93% confidence
Finding: The skill is written entirely as a Chinese-language creation workflow and describes its purpose in Chinese, but nowhere offers the user a language/locale choice or states that the skill is intentionally restricted to Chinese-only use. Under the policy, forcing a specific language without opt-in is a natural-language policy violation unless the locale constraint is explicit and justified.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill explicitly instructs the agent to scan the entire workspace for existing project folders and draft files, which can expose unrelated user data without an explicit, informed consent step. Because this happens at startup and is framed as default behavior, it creates a privacy boundary issue: the agent may inspect pre-existing files the user did not intend to share in the current interaction.

Natural-Language Policy Violations

Medium

Confidence: 96% confidence
Finding: The instruction hard-codes a binary 1:1 gender ratio and explicitly forbids changing existing characters' gender, forcing gender assignment based on quota rather than user intent or narrative need. In a creative-writing skill, this can systematically exclude non-binary identities and produce biased or inappropriate outputs, especially when users expect flexible or inclusive character design.

Natural-Language Policy Violations

Medium

Confidence: 96% confidence
Finding: This markdown skill file presents all operational instructions in Chinese and does not indicate that the user can choose another language or that the skill is intended only for a Chinese-language environment. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy violation.

Natural-Language Policy Violations

Medium

Confidence: 96% confidence
Finding: This markdown file contains natural-language instructions exclusively in Chinese, with no indication that the user may choose another language or locale. Under the policy rule, forcing a specific language without user opt-in is a natural-language policy violation.

Natural-Language Policy Violations

Medium

Confidence: 91% confidence
Finding: This file presents all operational instructions in Chinese and does not provide any user opt-in, alternative language option, or justification for a Chinese-only constraint. Under the policy, forcing a specific language without user choice is a natural-language policy violation.

Natural-Language Policy Violations

Medium

Confidence: 88% confidence
Finding: The file is written as a Chinese-only operating manual and presents the workflow as fixed in that language, without offering the user a language or locale choice. Under the policy, forcing a specific language without opt-in is a natural-language policy concern.

Missing User Warnings

Medium

Confidence: 96% confidence
Finding: The skill instructs the agent to create and later delete a draft file automatically after user confirmations, but it does not require any explicit user-facing disclosure or consent for those filesystem side effects. In an agent environment with write/delete capabilities, this can lead to unexpected persistence of user content and silent data removal, increasing the risk of privacy issues, workspace tampering, or accidental loss of material.

Missing User Warnings

Medium

Confidence: 98% confidence
Finding: This section directs the agent to create a project folder, generate multiple output files, save content, and delete the draft, again without mandating a clear warning or opt-in from the user. Because these are persistent filesystem operations, a user may trigger them unintentionally, and an over-privileged agent could write into sensitive workspace locations or remove intermediate data without the user realizing it.

Static analysis

No suspicious patterns detected.