Back to skill

Security audit

赛博朋克主题

Security checks for vulnerabilities and agentic risk

Overview

This skill openly installs a visual OpenClaw theme and changes live UI files by default, with backups and a skip-apply option, so it is not risk-free but the behavior is coherent with its purpose.

Install only if you want this skill to change the target workspace and the live OpenClaw Control UI. Use --skip-apply first if you want to stage files without modifying the live UI, and review backups before relying on the rollback path. Be aware that a fresh install without bundled/default assets may contact ClawHub to retrieve verified image assets.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (20)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 45)May include surrounding context.

md
`python3 scripts/install_cyberpunk_theme.py --workspace <target-workspace>`

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 49)May include surrounding context.

md
`python3 scripts/install_cyberpunk_theme.py --workspace <target-workspace>`

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 53)May include surrounding context.

md
`python3 scripts/install_cyberpunk_theme.py --workspace <target-workspace>`

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 89)May include surrounding context.

md
`python3 scripts/install_cyberpunk_theme.py --workspace <target-workspace>`

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
97% confidence
Finding

The skill clearly documents capabilities that read and write workspace files, execute Python scripts, modify a live Control UI, and may download assets from the network, yet it declares no explicit tool scope or permission boundaries. That mismatch is dangerous because an agent or user may invoke a skill with broad implicit authority without clear consent, making destructive or unexpected system changes easier.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The documentation states that installation will immediately apply the theme to the live OpenClaw control UI and generate reusable scripts in the target workspace, but it does not present a prominent warning about modifying a live environment. This is risky because users may run the command expecting a local theme install and instead alter production-like UI state, requiring rollback or causing disruption.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The display name and short description are presented in Chinese, while the file does not indicate that the skill is region-specific or that users can choose their preferred language. This creates a locale-policy concern because the skill appears to impose a specific language without opt-in or justification.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The default prompt advertises a broad invocation phrase that can trigger installation, repair, import, and asset-swapping behaviors without clearly constraining when or how the skill should be used. In an agent ecosystem, underspecified triggers increase the chance of accidental invocation, prompt-routing confusion, or user manipulation into applying theme changes in the wrong workspace or context.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The quota label normalization converts English labels like 'week'/'weekly' into the Chinese label '周' unconditionally. This is a natural-language locale choice embedded in code and does not offer a user language preference or document a justified region-specific constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The selector :root, :root[data-theme-mode="light"] assigns identical dark-theme variables and sets color-scheme:dark, overriding the declared light mode. This is a natural-language policy concern because the file comments and selectors indicate intentional theme enforcement without user opt-in or an explicit documented exception.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file's instructional content is presented only in Chinese, which can amount to forcing a specific language without user opt-in under the stated policy. There is no indication that this skill is region-specific or that users may choose another language.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The installer can fetch a remote ZIP bundle from clawhub.ai to restore missing theme assets, which expands the trust boundary beyond the locally bundled theme described by the skill. Although the code enforces a size limit and verifies SHA-256 hashes on decoded assets, this still introduces network dependency, external content retrieval, and a remotely triggered install path that could fail open operational expectations or increase attack surface if the download source, transport, or bundle composition is compromised.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/install_cyberpunk_theme.py (reported line 102)May include surrounding context.

python
return candidate

    try:
        npm_root = subprocess.run(
            ['npm', 'root', '-g'],
            check=True,
            capture_output=True,

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The natural-language instructions and operational guidance in this skill are predominantly Chinese, and the file does not indicate that the user can choose another language or that the skill is intended only for a Chinese-speaking audience. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

The aria-label text is fixed in Chinese ("赛博朋克主题封面"), which imposes a specific language in user-facing content. Because this file provides no opt-in, alternate locale, or justification that it is intended only for a Chinese-language context, it appears to violate the language/locale policy criterion.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The primary user-visible text includes Chinese strings such as "赛博朋克主题" and "可替换主助手 / Tool / 用户头像,以及聊天与梦境背景。" This fixes the asset to a specific language without indicating that users can choose another locale or that the asset is intentionally region-specific.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The image metadata uses a Chinese-only aria-label ("赛博朋克主题图标"), which imposes a specific language choice in user-facing text. This is a natural-language locale choice without any indication of user preference or documented regional scope.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

This markdown file gives users commands and manual steps to install or replace theme assets in a workspace, including rerunning an install script and hand-replacing target files. Although the behavior is expected for a theme-install skill, the document does not explicitly warn that existing files in those paths may be overwritten or modified.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

Beyond installing and swapping theme assets, the script reads IDENTITY.md and SOUL.md from the workspace and derives a nameplate string that is later injected into the live UI CSS. That behavior goes beyond the manifest’s stated scope of theme installation, compatibility repair, workspace import, and avatar/background swapping.

Content

No source excerpt is available for this finding.

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · scripts/install_cyberpunk_theme.py (reported line 263)May include surrounding context.

python
'dream_avatar',
        'dream_background',
    ):
        if getattr(args, attr) is not None:
            continue
        value = config_value(config, *CONFIG_KEY_ALIASES[attr])
        if isinstance(value, str) and value.strip():