Security audit

Wahoo

Security checks for vulnerabilities and agentic risk

Overview

This is a screenwriting guidance skill made of Markdown reference files, with usability caveats but no evidence of malware, data theft, destructive behavior, or hidden high-impact actions.

Reasonable to install for screenplay and story-development help. Be aware that broad creative-writing prompts may route to this skill, and some reference material is Chinese. Ask for concise explanations rather than internal chain-of-thought, and avoid sharing sensitive unpublished material unless you are comfortable with the platform's data handling.

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Vague Triggers

Medium

Confidence: 90% confidence
Finding: The trigger phrases are generic everyday requests such as 'I have an idea' and 'Help me perfect this story,' which can cause the skill to activate unintentionally during normal conversation. In an agent environment, broad activation increases prompt-routing ambiguity and can make the skill intercept requests outside a clearly scoped screenplay-writing context, leading to unintended behavior or policy bypass via misrouting.

Natural-Language Policy Violations

Medium

Confidence: 94% confidence
Finding: The file is entirely in Chinese and provides no language-selection mechanism or documented locale restriction, which can make the skill inaccessible or misleading for users, reviewers, and downstream agents expecting English or multilingual support. In an agent ecosystem, this can also hinder security review and safe operation because important instructions, limitations, or unsafe outputs may not be understood by the user or by operators monitoring the skill.

Natural-Language Policy Violations

Medium

Confidence: 91% confidence
Finding: The file content is entirely in Chinese and provides no indication that the skill is locale-restricted or that users can select another language. In a general-purpose agent skill, this can cause instruction mismatch, reduce transparency, and lead users or downstream systems to misunderstand outputs, which is a real policy/compliance and usability risk even if it is not a code-execution issue.

Natural-Language Policy Violations

Medium

Confidence: 92% confidence
Finding: The file is entirely in Chinese and does not provide a language choice, translation, or documented locale restriction. In a general-purpose agent skill, this can cause users or downstream reviewers to misunderstand instructions, miss safety-relevant constraints, or be unable to audit the content, which increases operational and security risk through opacity rather than direct exploitability.

Static analysis

No suspicious patterns detected.