Back to skill

Security audit

Seedream 5.0 Lite 图片生成与编辑

Security checks for vulnerabilities and agentic risk

Overview

This image-generation skill is mostly coherent, but it handles API keys and downloaded files in ways that could expose credentials or write outside the intended output folder.

Review this skill carefully before installing. Use it only with the official AI Hive endpoint, avoid custom --base-url or AI_HIVE_BASE_URL values unless you fully trust that endpoint, run it in a limited working account or container, and inspect output paths before opening downloaded files. A pinned dependency file and filename sanitization would materially reduce the risk.

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 (3)

T09 · Insecure Skill Coding Practices

Error
Location
scripts/imagegen.py:62
Finding

Bearer API Key Can Be Transmitted to an Arbitrary Configured Endpoint

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
scripts/imagegen.py:399
Finding

Server-Controlled Task Metadata Can Escape the Output Directory

Content
View full analysis
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
SKILL.md:72
Finding

Third-Party Dependency Is Installed Without Version or Integrity Pinning

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • 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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (9)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

The declared purpose says this is a Seedream 5.0 Lite image-generation skill, but the detected behavior reportedly includes broader AI Hive operations such as chat, video generation, user-info/wallet queries, model enumeration, and generic provider access. That mismatch is dangerous because operators may grant or invoke the skill under narrow expectations while the implementation can reach unrelated data and actions, expanding blast radius and enabling covert misuse.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
85% confidence
Finding

The skill declares no explicit tool scope or permissions while instructing execution of a Python script that can use shell, filesystem, environment variables, and network access. This creates an unnecessary trust gap: users and the orchestrator cannot tell from the manifest what capabilities will be exercised, increasing the chance of over-privileged execution or unintended data exposure.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest description says to use this skill when users search broad terms like '图片生成编辑' and '海报', which are generic image-generation and poster-related phrases rather than narrowly scoped invocations. Because no exclusion conditions or trigger constraints are provided, the skill may activate for ordinary image-editing requests that do not specifically call for Seedream 5.0 Lite or AI Hive.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The natural-language content of the skill is entirely in Chinese, including the operative description and usage guidance, while the file does not state that the skill is region-specific or that users may choose another language. This can violate language or locale policy when users have not opted into Chinese output or instructions.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The file embeds a broad AI Hive client with text chat, video generation, model listing, upload, and task features even though the declared skill is an image-only Seedream 5.0 Lite skill. This capability mismatch expands the attack surface and enables use of unintended models or modalities if the script is invoked outside the narrow wrapper, undermining least-privilege and user expectations.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The module docstring presents the artifact as a general-purpose AI capability tool rather than a narrowly scoped Seedream 5.0 Lite image skill. In context, that is a security-relevant integrity issue because it signals and normalizes broader behavior than the skill metadata promises, making misuse and hidden capability exposure more likely.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

This code file contains natural-language descriptions and CLI help text exclusively in Chinese, beginning with the module docstring and continuing throughout the script. Under the policy, forcing a specific language without opt-in is a natural-language policy violation unless the locale restriction is documented and justified, which is not present here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

This markdown file contains substantive skill-related documentation entirely in Chinese, including capability descriptions and operational notes. Under the policy rule, forcing a specific language without user opt-in can be a natural-language policy violation when no locale justification or language alternative is provided.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
94% confidence
Finding

The placeholder value 请替换为你的APIKey is written only in Chinese, which imposes a specific language in user-facing configuration content. There is no indication that the skill is region-specific or that users may choose another language, so this conflicts with the language/locale policy criteria.

Content

No source excerpt is available for this finding.