Back to skill

Security audit

Walmart 电商图片生成与编辑

Security checks for vulnerabilities and agentic risk

Overview

This skill mostly does the advertised image-generation work, but it can send API credentials and uploaded product media to any configured API origin without validation or clear warning.

Review before installing. Use only a trusted AI Hive endpoint, avoid setting --base-url or AI_HIVE_BASE_URL unless you control and trust the destination, and do not upload proprietary or pre-release product assets without approval. Consider using a throwaway or scoped API key and a virtual environment with pinned dependencies.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (2)

T09 · Insecure Skill Coding Practices

Error
Location
scripts/imagegen.py:57
Finding

Configurable API Origin Can Expose Credentials and Product Media to an Untrusted Server

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
SKILL.md:67
Finding

Unpinned Third-Party Dependency Installation

Content
View full analysis
``` ### Technical Analysis The installation instructions fetch `requests` without a pinned version or package hash. The effective code installed therefore depends on the latest package and transitive dependency versions available from the configured Python package index at installation time. The project does not include a lock file or hash-verified requirements file. This makes installations non-reproducible and prevents users from verifying that they received the dependency set reviewed with the Skill. This is a supply-chain hardening issue rather than evidence that the legitimate `requests` package is currently malicious. Exploitation depends on compromise or manipulation of the package source, package resolution process, or a future dependency release. ### Attack Path 1. An attacker compromises the configured package index, a relevant package release, the user's package-index configuration, or the dependency delivery path. 2. A user follows the documented `pip3 install requests` command. 3. `pip` resolves and installs the attacker-influenced package version or dependency set because no approved version or hash is enforced. 4. The installed code executes when `imagegen.py` imports `requests`. 5. Malicious dependency code then runs with the privileges of the user invoking the Skill. ### Impact Assessment A compromised dependency would execute inside the Skill's Python process with the invoking user's privileges. It could potentially access the API key, prompts, product media, downloaded outputs, environment variables, and other files available to that user. The exact ...[truncated 191 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Findings (8)

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The file implements generic text-chat and video-generation features even though the advertised skill is Walmart image generation/editing. In an agent setting, hidden or unnecessary capabilities are dangerous because they broaden what the skill can do with the configured API key, potentially enabling unreviewed model access, unexpected data transmission, and policy bypass through functionality outside the declared use case.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The file uses Chinese for all substantive changelog entries, which can constitute a language-policy issue when a skill forces a specific language without user opt-in. There is no indication that this skill is region-specific or that users can choose an alternate language.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding

The skill invokes Python scripts, installs dependencies, and references capabilities that imply shell, filesystem, environment, and network access, but it does not declare any tool-scope restrictions in the manifest. This creates an avoidable trust gap: an agent or reviewer cannot tell from the manifest what capabilities are intended, increasing the chance of overprivileged execution or unsafe reuse in broader contexts.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The description is very broad and encourages use across many Walmart and omnichannel image-generation scenarios without clear trigger conditions or boundaries. Broad invocation wording can cause accidental activation on loosely related tasks, leading to unintended execution of a skill that performs networked image processing on product assets.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The run instructions direct the user to initialize and execute an image-generation workflow against a fixed external endpoint, but they do not warn that supplied product reference images may be transmitted off-system. This can expose proprietary product imagery, packaging, labels, or pre-release retail assets to a third-party service without informed user consent or data-handling review.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The module docstring and subsequent user-facing guidance are written entirely in Chinese, and the CLI prompts/messages throughout the file assume Chinese-language operation. For a general-purpose skill, this imposes a specific language/locale on users without offering a choice or documenting a justified region-specific constraint.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill is presented as a narrowly scoped Walmart image-generation tool, but it also exposes additional operational capabilities such as task inspection and direct media upload. This scope mismatch can cause users or orchestrators to grant trust, permissions, or API access under a narrower assumption than the code actually warrants, increasing the attack surface and enabling unintended data access or service use.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The sample API key value contains Chinese-only instructional text ("请替换为你的APIKey"), which imposes a specific language in a user-facing configuration example. The file does not offer a language choice or explain a region-specific requirement, so this is a natural-language locale policy issue.

Content

No source excerpt is available for this finding.