Back to skill

Security audit

B站-关键词搜视频 | 青虎AI

Security checks for vulnerabilities and agentic risk

Overview

This Bilibili search skill is mostly purpose-aligned, but it gives the agent broad setup authority to install tools, change local configuration, and handle service tokens without enough user control.

Install only if you are comfortable letting this skill install or upgrade qhkit/Node and modify qhkit account configuration. Before use, require explicit confirmation for any install, PATH change, token configuration, site switch, or paid qhkit call, and prefer preinstalled/pinned qhkit tooling.

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
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (8)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The trigger language includes broad phrases like '同类需求' without clear exclusions, making accidental invocation more likely. In this skill, mis-triggering is more dangerous than normal because invocation can cascade into paid API calls, installation steps, and host configuration changes.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The documentation goes far beyond a simple search skill and instructs the agent to install Node/qhkit, modify PATH, configure tokens, and manage host tooling. This broadens the skill from data retrieval into host-environment bootstrapping, which materially increases the attack surface and can lead to unauthorized software changes or credential handling on the user's machine.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
90% confidence
Finding

The skill instructs use of npx @iqinghu/qhkit without pinning an exact package version. That allows whatever version is currently published or resolved at execution time to run in the host environment, expanding supply-chain risk for a skill whose core purpose is only Bilibili search.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
94% confidence
Finding

This duplicated finding points to the same persistence behavior: unpacking software into a user directory and exporting PATH for subsequent use. Even with checksum guidance, the core problem is that the skill authorizes durable host modification unrelated to the minimum function of searching Bilibili videos.

Content

Scanner excerpt · SKILL.md (reported line 90)May include surrounding context.

Linux 二进制安装(装到用户目录,无需 root;先校验官方 SHA256 再解包):

cd /tmp && curl -fsSLO https://nodejs.org/dist/v22.22.3/node-v22.22.3-linux-x64.tar.xz cd /tmp && curl -fsSL https://nodejs.org/dist/v22.22.3/SHASUMS256.txt | grep ' node-v22.22.3-linux-x64.tar.xz$' | sha256sum -c - mkdir -p "$HOME/.local/lib" && tar -xJf /tmp/node-v22.22.3-linux-x64.tar.xz -C "$HOME/.local/lib" export PATH="$HOME/.local/lib/node-v22.22.3-linux-x64/bin:$PATH"

text

Session Persistence

Medium
Category
Rogue Agent
Confidence
94% confidence
Finding

This duplicated finding points to the same persistence behavior: unpacking software into a user directory and exporting PATH for subsequent use. Even with checksum guidance, the core problem is that the skill authorizes durable host modification unrelated to the minimum function of searching Bilibili videos.

Content

Scanner excerpt · SKILL.md (reported line 90)May include surrounding context.

Linux 二进制安装(装到用户目录,无需 root;先校验官方 SHA256 再解包):

cd /tmp && curl -fsSLO https://nodejs.org/dist/v22.22.3/node-v22.22.3-linux-x64.tar.xz cd /tmp && curl -fsSL https://nodejs.org/dist/v22.22.3/SHASUMS256.txt | grep ' node-v22.22.3-linux-x64.tar.xz$' | sha256sum -c - mkdir -p "$HOME/.local/lib" && tar -xJf /tmp/node-v22.22.3-linux-x64.tar.xz -C "$HOME/.local/lib" export PATH="$HOME/.local/lib/node-v22.22.3-linux-x64/bin:$PATH"

text

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly tells the agent to inspect and modify credential/configuration state via qhkit config show, config set, and environment-variable guidance. For a video search skill, directing credential management is over-scoped and risks exposing, overwriting, or steering secrets and service endpoints in the host environment.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

L097 明确写明“site 必须是 cn”,并指示在检测到 intl 配置时进行覆盖,只有用户明确说明使用国际版账号时才保留。这属于默认强制特定区域/站点的语言策略,未先征求用户选择或提供对等选项。

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

L019 的“找合作视频”没有限定平台、动作或数据检索上下文,用户也可能是在寻求创意建议、商务合作方案或跨平台内容搜索,而非调用该技能。缺少上下文约束会提高误触发概率。

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.