Back to skill

Security audit

AKM Fashion

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent fashion-advice skill that asks for relevant body and wardrobe details, with a caution around an unpinned optional CLI install command.

Before installing, prefer the ClawHub listing or a pinned release/commit if available, and share only body or wardrobe details you are comfortable putting into a styling profile. The reviewed skill text itself is focused on fashion decisions and does not show hidden system changes or data-exfiltration behavior.

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:40
Finding

Unpinned Third-Party Installation from Mutable Sources

Content
View full analysis

Vulnerability Details

File Locations:

  • SKILL.md:40-43
  • README.md:52-54
  • README.zh-CN.md:52-54
  • clawhub-copy.md:11-15

Vulnerability Type: Unpinned npm CLI dependency and mutable GitHub installation source
Risk Level: Medium

Vulnerable Code

The affected documentation instructs users to execute the following installation command:

bash
npx skills add https://github.com/sirsws/akm --skill akm-fashion-strategist --full-depth

Technical Analysis

The installation command uses npx skills without specifying an exact package version. Consequently, npx can resolve and execute the version of the skills package available at installation time rather than a version reviewed alongside this project.

The GitHub source is also specified only as a repository URL:

text
https://github.com/sirsws/akm

It is not pinned to an immutable commit hash or verified release artifact. The effective content installed by --full-depth can therefore change after this audit without any modification to the audited files.

No malicious package, payload, or compromised upstream component was identified in the audited artifact. The vulnerability is the absence of dependency pinning and integrity controls, which creates a supply-chain trust boundary outside the reviewed project.

Attack Path

  1. An attacker compromises the npm package resolved by npx skills, the associated package publishing account, or the referenced GitHub repository.
  2. The attacker publishes a modified CLI version or changes content reachable through the repository's mutable default branch.
  3. A user follows the documented installation instructions.
  4. npx resolves and runs the unpinned third-party CLI, and the installer retrieves content from the mutable repository source.
  5. The altered installer or Skill content executes or is installed under the invoking user's account.
  6. The resulting behavior may differ f ...[truncated 929 chars]
Remediation
View remediation

Remediation Suggestions

  1. Pin the npm CLI to an exact reviewed version:

    bash
    npx --yes skills@<exact-reviewed-version> add ...
    
  2. Pin the GitHub source to a full immutable commit hash, if supported by the installer:

    bash
    npx --yes skills@<exact-reviewed-version> add https://github.com/sirsws/akm.git#<full-commit-hash> --skill akm-fashion-strategist --full-depth
    
  3. Prefer a signed, versioned release archive over a mutable branch or default repository reference.

  4. Publish a SHA-256 checksum for the expected release artifact and require users or the installer to verify it before installation.

  5. Use signed Git tags or provenance mechanisms such as Sigstore and verify the signature against a documented maintainer identity.

  6. Commit a lock file or installation manifest recording the exact CLI version, repository commit, and expected integrity digest.

  7. Update every duplicated installation command consistently in SKILL.md, both README variants, and the ClawHub listing.

  8. Document the permissions and files used by the installer so users can review the expected installation effects before running it.

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (36)

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · ELICITATION_PROMPT.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · ELICITATION_PROMPT.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · EXECUTION_PROMPT.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · EXECUTION_PROMPT.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · INPUT_TEMPLATE.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · INPUT_TEMPLATE.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · PUBLISHING_CHECKLIST.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · PUBLISHING_CHECKLIST.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · README.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · README.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · RECORD_TEMPLATE.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · RECORD_TEMPLATE.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · SAMPLE_RECORD.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · SAMPLE_RECORD.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · SKILL.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · agent-store-copy.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · clawhub-copy.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Hidden Instructions

High
Category
Prompt Injection
Confidence
60% confidence
Finding

Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.

Content

Scanner excerpt · clawhub-copy.zh-CN.md (reported line 1)May include surrounding context.

md
<!--
文件:clawhub-copy.zh-CN.md
核心功能:作为 Fashion skill 的中文长版发布页文案,供 ClawHub / OpenClaw 或同类技能市场直接复用。
输入:Fashion 分支的 AKM 定位、方法结构与使用边界。

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The comment states this is an '英文前置挖掘提示词' (English pre-elicitation prompt), which indicates the skill is designed to operate in English by default. This is a natural-language policy concern because it imposes a language choice without offering the user an explicit language or locale option.

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 explicitly identifies itself as a Chinese prompt and instructs behavior entirely in Chinese for the Fashion branch. Under the policy, forcing a specific language without user choice is a natural-language locale violation unless clearly documented as an opt-in or region-specific constraint.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The description states this is a Chinese execution prompt, which constrains the skill to a specific language. Under the policy, forcing a locale or language without explicit user choice is a natural-language policy violation unless the constraint is clearly documented and justified as region-specific.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
85% confidence
Finding

The README instructs users to run npx skills add against a remote GitHub repository without pinning a specific version, commit, or immutable release. This creates a supply-chain risk: if the upstream package, resolver behavior, or referenced repository content changes, users may install or execute unintended code.

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 metadata states that the skill provides an English sample output and public sample specifically in English. This is a language constraint expressed in natural language, but the file does not indicate user choice, opt-in, or a documented justification for enforcing English.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
95% confidence
Finding

The installation command uses npx skills without pinning a specific package version, which can cause users to execute an unexpected or newly published package version at install time. In a skill distribution context, this increases supply-chain risk because behavior may change over time or a compromised dependency could be pulled and executed on the user's machine.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.