Vultr Inference

Generate images and text using Vultr Inference API. Supports Flux image generation and various LLMs for text. Use when user wants to generate images, artwork...

Security Scan

The skill's requests and code match its stated purpose (calling Vultr Inference to generate images and chat completions); nothing in the files indicates hidden data exfiltration or unrelated privileges, though there are minor documentation inconsistencies to be aware of.

Purpose & Capability

Name/description (Vultr Inference) align with the code and SKILL.md: the script and examples call Vultr inference endpoints and download generated images. The only credential it needs is the Vultr API key stored in ~/.config/vultr/api_key, which is coherent with the stated purpose.

Instruction Scope

Overall instructions stay within scope (call inference endpoints, download images, list models). Minor inconsistencies: SKILL.md examples show using an environment variable ($VULTR_API_KEY) while the setup and included Python/script both read the API key from ~/.config/vultr/api_key; model naming in examples (e.g., flux.1-schnell, stable-diffusion-3.5-medium) differs from some defaults/options in the script (which uses short names like 'flux' and 'flux-pro'). These could cause confusion or errors but are not evidence of malice.

Install Mechanism

Instruction-only with a single helper script; there is no install spec or remote download. Nothing is written to system locations by an installer. Low-risk from installation perspective.

Credentials

The skill only accesses a single local file (~/.config/vultr/api_key) for the Vultr API key which is proportionate to its purpose. Caveats: the README also shows use of an env var ($VULTR_API_KEY) that is not declared as required; the skill expects the key stored in plain text, so users should protect file permissions and ensure the key has minimal privileges.

Persistence & Privilege

Skill does not request always:true, does not modify other skills or global settings, and does not persist beyond reading the API key and writing downloaded images to the current directory. Normal privileges for this kind of helper.

Assessment

This skill appears to do what it says: call Vultr Inference endpoints and download results. Before installing, consider: (1) the API key is read from ~/.config/vultr/api_key (plain text) — restrict that file's permissions (e.g., chmod 600) and ensure the key has only inference permissions; (2) SKILL.md examples also use $VULTR_API_KEY — decide which method you'll use and update docs if needed; (3) the script's default/allowed model names differ from some examples, so verify correct model IDs with your Vultr account if you see 400 errors; (4) generated images are written to the current directory — run the tool in a safe folder if you worry about overwriting files. If you want extra assurance, inspect the script locally (it is short and readable) before enabling autonomous invocation.

Generate images and text using Vultr's Inference API.

Setup

Uses the same API key as Vultr Cloud API. Store it at:

~/.config/vultr/api_key

Image Generation

Available Models

Model	Description
flux.1-dev	FLUX.1-dev - High quality
flux.1-schnell	FLUX.1-schnell - Fast generation
stable-diffusion-3.5-medium	SD 3.5 Medium - Balanced

Generate Image

curl -X POST "https://api.vultrinference.com/v1/images/generations" \
  -H "Authorization: Bearer $VULTR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "flux.1-schnell",
    "prompt": "a hedgehog eating a burger in Amsterdam",
    "n": 1,
    "size": "1024x1024"
  }'

Parameters

Parameter	Type	Description
model	string	flux.1-dev, flux.1-schnell, stable-diffusion-3.5-medium
prompt	string	Text description of image
n	int	Number of images (1-4)
size	string	256x256, 512x512, 1024x1024
response_format	string	url (default) or b64_json

Response

{
  "created": 1734567890,
  "data": [
    {
      "url": "https://ewr.vultrobjects.com/vultrinference-images/tmp_xxx.png"
    }
  ]
}

Text Generation (Chat Completions)

Available Models

• llama-3.1-405b-instruct - Meta Llama 3.1 405B
• llama-3.1-70b-instruct - Meta Llama 3.1 70B
• llama-3.1-8b-instruct - Meta Llama 3.1 8B
• mixtral-8x7b-32768 - Mixtral 8x7B
• qwen-2-72b-instruct - Qwen 2 72B

Chat Completion

curl -X POST "https://api.vultrinference.com/v1/chat/completions" \
  -H "Authorization: Bearer $VULTR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "llama-3.1-70b-instruct",
    "messages": [
      {"role": "user", "content": "Hello, how are you?"}
    ],
    "max_tokens": 100
  }'

Parameters

Parameter	Type	Description
model	string	Model ID from list above
messages	array	Chat messages with role and content
max_tokens	int	Maximum tokens to generate
temperature	float	Randomness (0-2, default 1)
stream	bool	Stream response (default false)

Python Example

import os
import requests

API_KEY = open(os.path.expanduser("~/.config/vultr/api_key")).read().strip()

# Generate image
response = requests.post(
    "https://api.vultrinference.com/v1/images/generations",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "model": "flux.1-schnell",
        "prompt": "a hedgehog eating a burger",
        "size": "512x512",
        "n": 1
    }
)

result = response.json()
image_url = result["data"][0]["url"]
print(f"Image URL: {image_url}")

# Download image
img_response = requests.get(image_url)
with open("generated_image.png", "wb") as f:
    f.write(img_response.content)

List Available Models

curl -s "https://api.vultrinference.com/v1/models" \
  -H "Authorization: Bearer $VULTR_API_KEY" | jq

Troubleshooting

401 Unauthorized

• Check API key is valid
• Ensure key has inference permissions

400 Bad Request

• Check model name is correct
• Check size is valid (256x256, 512x512, 1024x1024)
• Check prompt is not empty

Rate Limits

• Default: 60 requests per minute
• Contact support for higher limits