爆款视频多平台改版｜AI-HIVE

• Outcome: Pass
• Latest audit: Aug 25, 2026 · 12:33 PM UTC
• Version: 1.0.0

Overview

This skill is a disclosed AI-HIVE video workflow helper with expected API, media upload, local file, and ffmpeg behavior.

Before installing, be aware that using generation features requires an AI-HIVE API key and may upload selected media to AI-HIVE and incur costs. Use only authorized source material, review prompts and routing before generation, and avoid storing or sharing the API key in logs, screenshots, or repositories.

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection

Findings (6)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 87% confidence
• Finding: The skill advertises executable workflows that use environment variables, local file access, shell tools, and network calls, but it does not declare any corresponding permissions. This creates a transparency and containment problem: operators and users cannot accurately assess what the skill may access or invoke, increasing the chance of over-privileged execution or unsafe deployment assumptions.

Tp4

High

• Category: MCP Tool Poisoning
• Confidence: 92% confidence
• Finding: The documented behavior extends beyond the stated purpose into generic AI-HIVE chat/image generation, credential setup/storage, and general ffmpeg operations. This mismatch is dangerous because reviewers may approve the skill for a narrow video-adaptation use case while it actually enables broader actions, including credential handling and arbitrary media processing that expand the attack surface and may lead to data leakage or misuse.

Description-Behavior Mismatch

Medium

• Confidence: 94% confidence
• Finding: The file presents itself as a narrowly scoped multi-platform video reversion skill, but it also includes broad-purpose chat, image generation, model listing, upload, and task APIs. This scope expansion increases the attack surface and enables capabilities not justified by the manifest, undermining least-privilege expectations for users and reviewers.

Context-Inappropriate Capability

Medium

• Confidence: 90% confidence
• Finding: The skill exposes a user-info endpoint that retrieves account and wallet balance data, which is not necessary for generating platform-adapted videos. Unneeded access to account metadata creates privacy and reconnaissance risk, especially if the skill is invoked in shared or automated environments where such output may be logged or exposed.

Vague Triggers

Medium

• Confidence: 91% confidence
• Finding: The manifest enables implicit invocation without any visible trigger constraints, which can cause this skill to activate in broader contexts than the user intended. Because this skill can generate production-ready workflows and runnable AI-HIVE commands for multi-platform content adaptation, ambiguous activation increases the chance of unintended use for sensitive marketing, platform-policy-edge, or unauthorized replication requests.

Natural-Language Policy Violations

Medium

• Confidence: 83% confidence
• Finding: The default prompt hard-codes Chinese output without confirming user preference, which can override user expectations and reduce transparency about how the skill will behave. In a security context this is weaker than code execution issues, but it can still contribute to misleading or inappropriate operation if users receive output in an unexpected language and fail to properly review generated workflows or commands.

Static analysis

No suspicious patterns detected.