ClawHub

agnes-image-video

Security checks across malware telemetry and agentic risk

Overview

This skill is a straightforward Agnes AI image/video generation helper; it sends user prompts and image URLs to Agnes and saves returned media locally, which matches its stated purpose.

Install only if you are comfortable sending prompts, referenced image URLs, and Agnes account-linked API usage to Agnes AI. Use an environment variable or approved secret store for the API key, avoid sensitive or private images unless you intend to share them with the provider, and be aware that the helper script saves generated files locally and downloads provider-returned media URLs without extra validation.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (9)

Tainted flow: 'url' from httpx.post (line 54, network input) → httpx.get (network output)

Medium
Category
Data Flow
Content
if args.mode == "image":
        url = generate_image(api_key, args.prompt, args.size, args.seed, args.image_urls)
        # Download the image
        resp = httpx.get(url)
        ext = "jpg" if b"JPEG" in resp.content else "png"
        file_path = output_dir / f"image_{int(time.time())}.{ext}"
        file_path.write_bytes(resp.content)
Confidence
89% confidence
Finding
resp = httpx.get(url)

Tainted flow: 'video_url' from httpx.get (line 104, network input) → httpx.get (network output)

Medium
Category
Data Flow
Content
)
        # Download the video
        print("正在下载视频...")
        resp = httpx.get(video_url)
        file_path = output_dir / f"video_{int(time.time())}.mp4"
        file_path.write_bytes(resp.content)
        print(f"视频已保存: {file_path}")
Confidence
90% confidence
Finding
resp = httpx.get(video_url)

Lp3

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding
The skill documents code-like capabilities that use environment variables, network access, and local file output, but it does not declare corresponding permissions or constraints. This creates a transparency and governance gap: users and the host agent may not realize the skill can exfiltrate prompts/media to a third party and write generated outputs locally.

Vague Triggers

Medium
Confidence
86% confidence
Finding
The trigger scope is broad enough to activate on generic image/video requests, which can cause unintentional routing of user content to this third-party service. Overbroad activation increases the chance that sensitive prompts, images, or attachments are sent externally without the user specifically choosing this provider.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
The skill instructs the agent to send prompts, image URLs, and API credentials to an external API but does not provide an explicit privacy warning or informed-consent step. This is dangerous because users may unknowingly disclose sensitive text, media, or secrets to a third-party processor, creating privacy, compliance, and credential-handling risks.

External Transmission

Medium
Category
Data Exfiltration
Content
```python
import httpx

response = httpx.post(
    "https://apihub.agnes-ai.com/v1/images/generations",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
Confidence
91% confidence
Finding
httpx.post( "https://

External Transmission

Medium
Category
Data Exfiltration
Content
图生图**必须**在 `tags` 中添加 `"img2img"`，否则模型会当作文生图处理：

```python
response = httpx.post(
    "https://apihub.agnes-ai.com/v1/images/generations",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
Confidence
92% confidence
Finding
httpx.post( "https://

External Transmission

Medium
Category
Data Exfiltration
Content
在 `extra_body.image` 中传入多个图像 URL：

```python
response = httpx.post(
    "https://apihub.agnes-ai.com/v1/images/generations",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
Confidence
90% confidence
Finding
httpx.post( "https://

External Transmission

Medium
Category
Data Exfiltration
Content
import time

# Step 1: 创建视频任务
create_resp = httpx.post(
    "https://apihub.agnes-ai.com/v1/videos",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    json={
Confidence
90% confidence
Finding
httpx.post( "https://

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal