Back to skill

Security audit

探索式测试

Security checks for vulnerabilities and agentic risk

Overview

This is mostly a coherent QA exploratory-testing skill, but it recommends an unpinned npx install of a larger third-party skill set that could change what gets executed or installed.

Review the referenced Kokxi/qa-test-skills package or repository before running the npx install command, prefer a pinned commit or version with integrity checks, and avoid using real customer, payment, identity, or production data while generating exploratory-test outputs.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:8
Finding
Unpinned Third-Party Installation Through npx## Vulnerability Details **File Location**: `SKILL.md`, lines 8 and 54 **Vulnerability Type**: Unpinned third-party package and mutable external skill dependency **Risk Level**: Medium **Complete Code Snippets**: Line 8: ```text 本技能属于 QA Test Skills 技能集(49 个技能之一),完整工作流体验需安装全套:npx skills add Kokxi/qa-test-skills ``` Line 54: ```text > ⚠️ 本技能单独使用效果有限,建议配合完整技能集(12 步工作流)使用。安装:npx skills add Kokxi/qa-test-skills ``` ### Technical Analysis The Skill recommends invoking `npx` to install content from `Kokxi/qa-test-skills`. The command does not pin the `skills` npm package to an exact reviewed version, and the external skill collection is not pinned to an immutable release or commit. If `skills` is not already installed locally, `npx` may retrieve and execute package code from the npm registry. The resulting installation can also incorporate mutable content from the referenced external repository. The project supplies no lockfile, checksum, signature, or vendored source that would allow users to verify that the downloaded components match an audited version. This is a supply-chain weakness rather than evidence that the currently referenced package or repository is malicious. Exploitation depends on compromise, replacement, or malicious modification of one of the external components. ### Attack Path 1. An attacker compromises the npm package, its publisher account, the external repository, or another relevant distribution component. 2. The attacker publishes malicious package code or modifies the externally hosted Skill content. 3. A user follows the recommendation and runs `npx skills add Kokxi/qa-test-skills`. 4. `npx` retrieves and potentially executes the unpinned package under the invoking user's privileges. 5. The installer retrieves or installs attacker-controlled Skill content. 6. That code or content can operate within the permissions available to the process or subsequently influe ...[truncated 673 chars]
Remediation
## Remediation Suggestions - Pin the `skills` npm package to an exact, reviewed version rather than allowing `npx` to resolve a mutable version. - Pin `Kokxi/qa-test-skills` to a reviewed commit hash or immutable signed release. - Publish and verify cryptographic checksums or signatures for all downloaded artifacts. - Prefer a locally vendored and audited installer where practical. - Use a package lockfile or equivalent integrity metadata for executable dependencies. - Warn users explicitly that the command may download and execute third-party code. - Require explicit user approval before running the installation command. - Run installation in a sandbox or least-privileged environment without production credentials. - Re-audit external Skill content whenever its pinned version is updated.
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (5)

Natural-Language Policy Violations

High
Confidence
91% confidence
Finding
L004 将 displayName 固定为中文“探索式测试”,且整份技能文档均以中文撰写,没有说明语言可由用户选择,也没有给出这是区域性/合规性要求的理由。根据语言/locale 政策,强制特定语言而无用户选择属于自然语言策略违规。

Rp1

Medium
Category
MCP Rug Pull
Confidence
70% confidence
Finding
MCP server references in the skill manifest without version pinning are a rug-pull risk.

Rp1

Medium
Category
MCP Rug Pull
Confidence
95% confidence
Finding
The skill instructs users to install a package via `npx skills add Kokxi/qa-test-skills` without pinning an exact version. Because `npx` resolves the latest available package by default, a compromised upstream release or dependency update could cause users to execute unintended code during installation.

Vague Triggers

Medium
Confidence
94% confidence
Finding
L010 的 when_to_use 同时列出“到处点一点”“自由测试”“场景发现”等表述,其中“到处点一点”尤其接近日常随口表达,缺少明确上下文约束,容易与非该技能意图的普通请求发生重叠。该行也未提供排除条件或负例,触发边界不够清晰。

Rp1

Medium
Category
MCP Rug Pull
Confidence
95% confidence
Finding
This line repeats an unpinned `npx skills add` installation command. Recommending execution of an unversioned remote package increases supply-chain risk because the behavior can change over time or be hijacked by a malicious publication/update.

Static analysis

No suspicious patterns detected.