Back to skill

Security audit

Ai Intelligent Image Generation

Security checks for vulnerabilities and agentic risk

Overview

The skill has a coherent image-generation purpose, but its install instructions fetch and run unpinned remote code that was not included in the reviewed package.

Review the remote GitHub repository and dependency list before installing, preferably pin a known commit and use locked dependency hashes. Avoid submitting confidential prompts or private images unless you understand which external AI services receive them and their data policies.

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:38
Finding
Mutable Remote Repository and Unverified Dependencies Are Retrieved and Executed## Vulnerability Details **File Location**: `SKILL.md`, lines 38–41 **Vulnerability Type**: Remote payload retrieval and insecure dependency execution **Risk Level**: High ```bash git clone https://github.com/openclaw-skills/ai-intelligent-image-generation cd ai-intelligent-image-generation pip install -r requirements.txt python app.py ``` ### Technical Analysis The installation instructions clone an external Git repository without pinning it to a reviewed commit hash or signed release. They then install packages from a remote-controlled `requirements.txt` and execute the repository's `app.py`. Neither `requirements.txt` nor `app.py` is present in the audited project, which contains only `SKILL.md`. Consequently, the effective dependency set and executed application cannot be inspected or verified as part of this audit. The remote repository's default branch can change after review, making the executed payload mutable. In addition, `pip install` may execute package build or installation logic, so an unsafe or compromised dependency could run code even before `app.py` starts. ### Attack Path 1. An attacker compromises the referenced repository, gains control of an upstream dependency, or causes a malicious dependency to be included in `requirements.txt`. 2. The attacker modifies the remote application, dependency declarations, or package installation logic. 3. A user or agent follows the documented installation commands. 4. `git clone` retrieves the attacker-controlled content because no immutable commit is specified. 5. `pip install -r requirements.txt` downloads and may execute unverified dependency installation code. 6. `python app.py` directly executes the retrieved application with the invoking user's privileges. ### Impact Assessment Successful exploitation could provide arbitrary code execution under the account running the commands. The accessible scope could include that user's files, environment variable ...[truncated 399 chars]
Remediation
## Remediation Suggestions 1. Include the implementation and dependency manifests in the reviewed skill package so that all executed content is available for audit. 2. If remote retrieval is necessary, pin the repository to a specific verified commit hash rather than cloning a mutable default branch. 3. Use signed releases or verify downloaded content against a trusted cryptographic digest before installation or execution. 4. Pin every Python dependency to an exact version and use a lock file containing hashes, such as hash-locked requirements installed with `pip --require-hashes`. 5. Review direct and transitive dependencies, and obtain packages only from explicitly trusted indexes. 6. Avoid executing dependency build scripts where feasible; prefer reviewed, reproducible artifacts. 7. Run installation and application startup in an isolated, least-privileged environment with restricted filesystem access, sanitized credentials, and controlled network egress. 8. Document the expected files, verified revision, dependency versions, and integrity checks directly in `SKILL.md`.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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)

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill advertises integration with external AI image services such as DALL-E API, but the description does not warn users that prompts and uploaded images may be transmitted to third-party providers. This creates a real privacy and data-handling risk because users may unknowingly submit sensitive text or images to external services with separate retention, logging, or training policies.

Natural-Language Policy Violations

Low
Confidence
79% confidence
Finding
The natural-language content of the skill, including its name/description sections and setup guidance, is presented exclusively in Chinese. This can constitute a language policy issue when the skill does not offer an alternative language, user opt-in, or a documented reason for the locale restriction.

Static analysis

No suspicious patterns detected.