Back to skill

Security audit

AI大模型专家|Seedance2.0 视频生成与编辑

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a disclosed AI-HIVE video-generation client, but it has review-worthy credential and invocation-scope risks.

Install only if you intend to use AI-HIVE/Seedance and are comfortable storing an API key locally and uploading selected media to the service. Avoid custom `--base-url` or `AI_HIVE_BASE_URL` values unless you fully trust the endpoint, and prefer a pinned dependency install instead of the documented unpinned `requests` command. Disable or avoid implicit invocation for ambiguous content-generation requests.

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)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:204
Finding

Unpinned Third-Party Dependency Installation

Content
View full analysis
Remediation
View remediation
--hash=sha256: ``` 3. Install with hash enforcement: ```bash python3 -m pip install --require-hashes -r requirements.txt ``` 4. Record the approved Python and dependency versions so installations are reproducible. 5. Use an explicitly trusted package index or controlled internal mirror. 6. Add automated dependency vulnerability and update review rather than silently resolving the latest release. 7. Replace the unversioned installation command in both the documentation and runtime error message with the hardened installation process. ]]>

T09 · Insecure Skill Coding Practices

Warning
Location
scripts/videogen.py:90
Finding

Bearer API Key Can Be Sent to an Arbitrary Configured Endpoint

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (9)

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding

The skill describes and enables powerful capabilities including shell execution, file read/write, environment access, and outbound network access, but it does not declare any explicit tool scope or permissions boundary. In an agent ecosystem, this increases the chance the skill is invoked with broader privileges than users or orchestrators expect, enabling unauthorized file access, secret exposure, or arbitrary network/API actions.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The activation description is very broad and covers generic AI, ecommerce, advertising, image, and video workflows, making the skill eligible for many loosely related requests. Over-broad matching can cause unintended invocation of a networked, file-writing skill in contexts where the user did not specifically request this platform, increasing the risk of surprise external API calls, uploads, and local side effects.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The original search-intent section contains catch-all language aimed at competitors, migration scenarios, broad ecommerce platforms, and many generic content-production needs. This can intentionally or accidentally maximize activation frequency, causing the agent to route unrelated requests into a skill that performs uploads, task submission, polling, and downloads against a third-party service.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The skill enables allow_implicit_invocation without any visible trigger constraints, approval gates, or narrowing conditions in the manifest. Because this skill can submit external AI-HIVE jobs, upload user-provided media, poll task status, and download outputs, implicit invocation could cause unintended external actions, data transfer, and cost-incurring operations from ambiguous user requests.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The top-level docstring says this file is a '通用 AI 能力调用工具' that wraps text chat, image generation, video generation, model query, media upload, and task polling. However, the manifest presents this skill as a Seedance2.0 video generation/editing skill, and the actual entrypoint used by this file (skill_main) is specialized around Seedance video workflows rather than a general-purpose multimodal client.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

This code file contains prominent natural-language descriptions and user-facing output entirely in Chinese, beginning with the module docstring and continuing through CLI guidance. Under the policy rule for language/locale violations, the skill does not offer the user any language choice or opt-in despite being a general-purpose tool.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The metadata sets the language to 'zh-CN', and the document is written as a Chinese-only skill without indicating that users may choose another language. Under the stated policy, forcing a specific language or locale without opt-in can be a natural-language policy violation unless the constraint is clearly justified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

The example API key value uses Chinese text ('请替换为你的APIKey') as the only user-facing instruction in the file. For a general configuration example, this imposes a specific language without offering a choice or documenting that the skill is intended only for Chinese-speaking users.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

SKILL_CONFIG.editing_note states the skill does not claim a native edit endpoint and only supports regeneration constrained by reference materials. Yet the same config keywords and example market the skill for '视频编辑', '视频重绘', '视频续写', and skill_generate forwards --video, --image, --audio, --first-frame, and --last-frame into a generic video generation endpoint that functionally presents editing-like behavior. This is not a code-execution flaw, but the inline intent statement is in tension with how the skill is positioned and exposed in-code.

Content

No source excerpt is available for this finding.