Back to skill

Security audit

Ai Intelligent 3d Model Generation

Security checks for vulnerabilities and agentic risk

Overview

The skill is a thin wrapper that tells users to download and run unaudited remote Python code for 3D generation.

Review the referenced repository and dependencies before installing, pin a known commit, and run it in an isolated environment without sensitive files or credentials. Treat this as a Review item, not proof of malicious intent.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:39
Finding

Unpinned Remote Repository and Dependency Execution

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 39–42
Vulnerability Type: Remote payload retrieval and insecure dependency execution
Risk Level: High

Vulnerable Code:

bash
git clone https://github.com/openclaw-skills/ai-intelligent-3d-model-generation
cd ai-intelligent-3d-model-generation
pip install -r requirements.txt
python app.py

Technical Analysis

The documented installation process clones a mutable remote repository without pinning it to a reviewed commit, tag digest, or cryptographically verified release. It then installs dependencies from an unavailable requirements.txt and executes an unavailable app.py.

Because neither the remote application nor its dependency manifest is included in the audited artifact, their behavior cannot be verified. The effective payload may change after this Skill has been reviewed. Moreover, pip install can execute package build or installation logic, while python app.py directly runs the downloaded application with the invoking user's permissions.

The artifact supplies no commit hash, integrity checksum, signed release verification, dependency lockfile, package hashes, or source code corresponding to its advertised 3D-generation features.

Attack Path

  1. A user or AI agent follows the installation commands in SKILL.md.
  2. Git clones the repository's current default branch rather than a fixed, previously reviewed revision.
  3. An attacker who compromises the repository or a referenced dependency modifies requirements.txt, application files, or package installation behavior.
  4. pip install -r requirements.txt retrieves and installs the attacker-controlled component; package build or installation hooks may execute code.
  5. python app.py executes the mutable remote application directly.
  6. The payload operates with the privileges and environment access of the user running the commands.

Impact Assessment

Successful e ...[truncated 654 chars]

Remediation
View remediation

Remediation Suggestions

  1. Include the complete reviewed implementation and dependency declarations within the Skill package so its executable behavior can be audited.
  2. If remote retrieval is unavoidable, pin the repository to a specific verified commit hash rather than cloning a mutable default branch.
  3. Distribute signed releases and verify signatures or cryptographic checksums before installation or execution.
  4. Replace the unverified dependency manifest with a lockfile containing exact versions and package hashes, and install with hash verification enabled.
  5. Review all direct and transitive dependencies for provenance, typosquatting, dependency confusion, known vulnerabilities, and unsafe installation hooks.
  6. Do not automatically execute newly downloaded code. Require a separate verification and explicit approval step before running app.py.
  7. Perform installation and execution in a least-privileged, isolated environment without sensitive credentials, host filesystem access, or unnecessary network access.
  8. Keep the implementation, version metadata, dependency lockfile, and reviewed commit synchronized for each published Skill release.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest description and main documentation present the skill in Chinese only, and there is no indication that users can opt into another language or that the skill is intended solely for a Chinese-speaking or region-specific audience. This creates a natural-language policy concern because the skill effectively forces a specific language without user choice.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The description is written entirely in Chinese ('AI 3D模型生成 - 文生3D、图片转3D模型') with no indication that the skill supports other languages or that Chinese is required for a region-specific purpose. This can constitute a language/locale policy issue because it implicitly constrains presentation to a specific language without user opt-in.

Content

No source excerpt is available for this finding.