Back to skill

Security audit

抖音-达人主页视频 | 青虎AI

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent Douyin video-list retrieval integration, with expected external API use and paid-credit handling disclosed, though users should be aware it may install qhkit/Node and send profile URLs to Qinghu.

Before installing, confirm you are comfortable using Qinghu/qhkit, providing a Qinghu API token, sending Douyin profile URLs to that service, and allowing npm/Node tooling to be installed or updated on your machine. For bulk video pulls, review the credit estimate first because calls are paid.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (7)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The top-level description uses broad trigger language such as pulling all videos or analyzing a benchmark account, which can overlap with common user requests and cause over-invocation of this external-data skill. Overly broad activation increases the chance of unintended third-party data access, unwanted charges, or disclosure of user-supplied URLs to the external service.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill tells the agent to transmit a user-supplied Douyin profile URL to an external service and retrieve third-party account data, but it omits an explicit privacy and data-handling warning. Users may not realize their supplied links and queried account data are being sent to a separate provider, creating consent and privacy-risk issues.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The '何时触发' section provides only positive examples and no scope boundaries or negative examples, so an agent may match loosely related requests and invoke the skill when it should not. In this skill, mistaken invocation can send third-party profile URLs to an external provider and incur paid API calls.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
92% confidence
Finding

The skill instructs fallback execution via npx @iqinghu/qhkit without pinning an exact package version. That makes runtime behavior depend on the latest package published at execution time, which introduces supply-chain risk and undermines reproducibility if the package is updated maliciously or incompatibly.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
84% confidence
Finding

The duplicated finding reflects the same persistence concern: the skill prescribes unpacking binaries into the home directory and exporting a modified PATH for later use. In agent contexts, instructions that bootstrap tooling onto the host can become dangerous because they change the execution environment beyond the immediate query and may be performed without sufficiently explicit user consent.

Content

Scanner excerpt · SKILL.md (reported line 89)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
84% confidence
Finding

The duplicated finding reflects the same persistence concern: the skill prescribes unpacking binaries into the home directory and exporting a modified PATH for later use. In agent contexts, instructions that bootstrap tooling onto the host can become dangerous because they change the execution environment beyond the immediate query and may be performed without sufficiently explicit user consent.

Content

Scanner excerpt · SKILL.md (reported line 89)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

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
76% confidence
Finding

The guidance mandates the cn site and tells the agent to direct users to the domestic iqinghu.com flow by default, only preserving intl when the user explicitly states they use AutoAGC. This is a locale/site constraint imposed without initial user choice, which may conflict with language or locale preference policy.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.