Back to skill

Security audit

Multi Agent Dev

Security checks for vulnerabilities and agentic risk

Overview

This skill is a coherent multi-agent coding workflow, but it can route very broad coding requests into external iterative agents with limited confirmation and an unpinned global dependency.

Review this before installing if you do not want generic coding requests to launch external iterative agents. Pin and verify the Ralph Wiggum npm package, avoid elevated npm installs, require confirmation before spawning agents or creating directories, and use the skill only inside intended project workspaces.

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
README.md:65
Finding

Unpinned Global Installation of a Third-Party npm Package

Content
View full analysis

Vulnerability Details

File Location: README.md, line 65
Vulnerability Type: Unpinned third-party dependency installed globally
Risk Level: Medium

Vulnerable Code

bash
npm install -g @th0rgal/ralph-wiggum

Technical Analysis

The documented installation command retrieves the current version of @th0rgal/ralph-wiggum from the npm registry and installs it globally. It does not pin an exact reviewed version or verify the downloaded package with a cryptographic integrity value.

Consequently, the package installed by a user may differ from the version that was reviewed when this documentation was published. npm installation can also execute package lifecycle scripts. If the package, its maintainer account, its dependencies, or its publication channel is compromised, malicious installation logic could run with the permissions of the user invoking npm.

The linked source repository does not by itself cryptographically establish that the npm artifact being installed corresponds to reviewed source code.

Attack Path

  1. An attacker compromises the npm package maintainer account, a transitive dependency, or another part of the package publication channel.
  2. The attacker publishes a malicious release under the same package name and makes it the version selected by npm.
  3. A user follows the prerequisite command in README.md.
  4. npm downloads the unreviewed release and may execute its lifecycle scripts during global installation.
  5. Malicious code runs under the invoking user's account and can access resources available to that account.

Impact Assessment

Successful exploitation could permit arbitrary code execution with the invoking user's privileges. The attacker could read or modify user-accessible files, obtain credentials exposed to the process or stored in accessible configuration files, tamper with development tooling, and alter globally installed npm components.

The command does not explicitly require administrative p ...[truncated 223 chars]

Remediation
View remediation

Remediation Suggestions

  1. Pin the dependency to an exact version that has been reviewed:

    bash
    npm install -g @th0rgal/ralph-wiggum@<reviewed-exact-version>
    
  2. Prefer a project-local dependency recorded in package.json and a committed lockfile rather than a global installation.

  3. Document how users can verify package provenance and cryptographic integrity before installation.

  4. Use npm provenance attestations or an equivalent trusted artifact-verification mechanism where available.

  5. Recommend installing without lifecycle scripts when the package is compatible with that restriction:

    bash
    npm install --ignore-scripts @th0rgal/ralph-wiggum@<reviewed-exact-version>
    
  6. Avoid recommending sudo or other elevated execution for npm installation.

  7. Periodically re-review the pinned package release and its transitive dependency tree before updating the documented version.

Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Vague Triggers

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The trigger phrases are extremely broad, including generic terms such as '开发', '研发', and '编码', which can cause this skill to activate for many ordinary requests unrelated to this workflow. In a skill that can invoke external agents and perform iterative development actions, overbroad activation increases the chance of unintended execution and unsafe file or command operations without clear user intent.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The README content and usage instructions are written entirely in Chinese, with no indication that other languages are supported or that the user can choose their preferred language. This creates a natural-language locale policy concern because the skill documentation effectively imposes a specific language without opt-in or documented justification.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

Natural-language instructions, headings, triggers, and examples are presented entirely in Chinese, which effectively forces a specific language for use of the skill. There is no opt-in, bilingual alternative, or documented justification that this skill is intended only for a Chinese-language environment.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The exception handling guidance states that if the working directory does not exist, the skill should automatically create it or prompt the user, which normalizes making filesystem changes as part of recovery. Automatic directory creation can lead to unexpected writes in sensitive paths, typo-squatted locations, or user confusion about what the agent changed on disk.

Content

No source excerpt is available for this finding.

Skill Enumeration

Medium
Category
Agent Snooping
Confidence
80% confidence
Finding

Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.

Content

Scanner excerpt · SKILL.md (reported line 211)May include surrounding context.

md
### 1. 任务描述要具体
❌ 模糊:「帮我优化一下 skill」
✅ 具体:「在 `~/.openclaw/skills/xxx/SKILL.md` 末尾增加「使用示例」章节,包含2个代码示例」

### 2. 设置合理的 max-iterations
- 简单任务:3-5次

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

This markdown file contains user-facing natural language entries in Chinese, but there is no indication that the skill is region-specific or that users can opt into this language. Under the language/locale policy rule, forcing a specific language without user choice can be a policy violation.

Content

No source excerpt is available for this finding.