Back to skill

Security audit

AliExpress 速卖通 电商图片生成与编辑

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a disclosed AliExpress image-generation wrapper, but its upload command can transmit any readable local file to a third-party service without image-only validation or confirmation.

Review before installing. Use this only with files you intentionally want to send to AI Hive, preferably product images in a dedicated folder. Do not let untrusted prompts choose --image or --file paths, and consider pinning dependencies and using a virtual environment. The API key persistence is normal for this tool, but it stores a credential under ~/.ai-hive/config.json.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • 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
Findings (2)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:69
Finding

Unpinned Runtime Dependency Creates a Mutable Supply-Chain Trust Boundary

Content
View full analysis
Remediation
View remediation
``` 2. Generate and verify hashes for the package and all transitive dependencies. 3. Install dependencies with hash enforcement: ```bash python3 -m pip install --require-hashes -r requirements.txt ``` 4. Use a project-specific virtual environment instead of modifying the global Python environment. 5. Document the expected package index and discourage installation from untrusted mirrors. 6. Add automated dependency vulnerability and update checks while retaining explicit review before version changes. ]]>

T05 · Unauthorized Access and Privilege Escalation

Warning
Location
scripts/imagegen.py:266
Finding

Image Upload Interface Accepts Arbitrary Readable Local Files

Content
View full analysis
" raise SystemExit( f"Object-storage upload failed ({oss_resp.status_code}): {detail}" ) print(f"[3/3] Confirming upload completion...") result = client.complete_upload(media_id) print(f"[ok] mediaId = {media_id}") return media_id ``` The Skill exposes this routine through an image-labe ...[truncated 2491 chars]
Remediation
View remediation
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (9)

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The file is marketed as an AliExpress image-generation/editing skill, but the implementation is a general-purpose AI Hive client with chat, video, model enumeration, account lookup, upload, and task management capabilities. This violates least-privilege and creates capability overreach: a caller expecting a narrowly scoped ecommerce-image tool can instead access unrelated AI functions and account-linked operations.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

Video-generation support materially exceeds the stated image-generation/editing purpose and introduces additional data-handling paths for uploaded images, videos, and audio. This broader media capability increases attack surface and may let users generate or upload content outside the controls expected for an ecommerce image-editing skill.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding

The skill invokes shell commands, installs packages, performs network access, and reads/writes files, but it does not declare any tool scope or permissions boundaries. That creates an implicit overprivilege problem: an agent may grant broad execution capabilities to a skill that appears to be simple image-generation guidance, increasing the chance of unintended code execution, filesystem access, or data exfiltration if the script or its inputs are abused.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The skill's visible instructions and operational guidance are predominantly written in Chinese, while the file does not offer an explicit language choice or opt-in for users who may need another language. 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.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The module documentation presents the code as a generic AI capability tool, directly contradicting the advertised AliExpress-specific image skill identity. This kind of scope misrepresentation is dangerous because reviewers and orchestrators may grant trust or permissions based on the manifest while the code actually supports much broader operations.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This code file contains natural-language descriptions and prompts that assume Chinese as the interaction language, including the module docstring and later CLI help/output strings. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly justified, which is not present here.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

Fetching user account information and wallet balance is not necessary for producing or editing product images, and it exposes sensitive account context through an unrelated skill. In a delegated or agent-driven environment, this can leak billing or identity data beyond user expectations.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The chat_text capability enables arbitrary text or multimodal prompting unrelated to product image generation, expanding the skill into a general LLM interface. In a skill ecosystem, this can be abused for unintended data processing, prompt exfiltration, or bypassing governance that assumes the skill is image-only.

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 includes Chinese text ("请替换为你的APIKey"), which imposes a specific language in user-facing configuration content. The file does not provide an alternative language option or explain that the skill is intentionally limited to Chinese-speaking users.

Content

No source excerpt is available for this finding.