100Critical
The skill is vulnerable to command injection and prompt injection
Behavior Analysis
Mismatch detected — The skill's stated purpose heavily emphasizes 'true privacy' and 'zero data retention'. While the skill provides options to enable Venice AI's claimed privacy features (e.g., E2EE, private models), the skill itself is a client that sends user data (prompts, files) to the remote Venice AI service. The privacy guarantees are entirely dependent on the Venice AI service's implementation and trustworthiness, not on the local execution of the skill.
Claims to do
Venice AI — Complete Platform Skill: **Venice AI is the AI platform for people who don't want Big Tech watching over their shoulder.** Unlike OpenAI, Anthropic, and Google — where every prompt is logged, analyzed, and potentially used to train future models — Venice offers **true privacy** with zero data retention on private models. Your conversations stay yours.
Actually does
The skill executes various Python scripts (`venice.py`, `venice-image.py`, etc.) using `python3`. It requires a `VENICE_API_KEY` to authenticate and interact with the Venice AI platform's API. These scripts send user prompts, local files (images, audio, video), and specified URLs to the Venice AI service for tasks like text generation, vision analysis, embeddings, TTS, speech-to-text, image/video/music generation, and editing.
Skill Info
Namejonisjongithub/venice-ai
Registryclawhub
Version2.1.1
PURLpkg:clawhub/jonisjongithub/venice-ai@2.1.1
Stars3
Downloads2,692
Installs9
Source
SHA-2562dd1fe9468ea...
SHA-115a441729526...
MD569c8144fe138...
TLSH7eb285537c90...
ssdeep384:I3Q+YX6J...
Size23,601 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
ClawHubDocs
openclaw skills install jonisjongithub/venice-aiAssessments (3)
AI Agent Traps ↗Prompt Injection Detectedcriticalprompt injection
DeBERTa classifier detected prompt injection (confidence: 0.98)
98% confidence
Command Execution & Arbitrary File/URL Accesshighcommand execution
The skill executes local Python scripts with user-provided arguments, including file paths and URLs. This creates a risk of command injection if inputs are not properly sanitized, allowing arbitrary command execution, broad read/write access to the local filesystem, and potential Server-Side Request Forgery (SSRF) via URL parameters.
90% confidenceline 124
python3 {baseDir}/scripts/venice.py chat "..."
python3 {baseDir}/scripts/venice.py analyze photo.jpg
python3 {baseDir}/scripts/venice.py analyze "https://example.com/image.jpg"
python3 {baseDir}/scripts/venice.py tts "..." --output /tmp/speech.mp3API Key Handling and Potential Exfiltrationmediumcredential theft
The skill requires a `VENICE_API_KEY` to be set as an environment variable. While not hardcoded, the Python scripts that interact with the Venice API will access this key, posing a risk of exfiltration if the scripts are compromised or vulnerable to command injection.
80% confidenceline 14
requires: { "bins": ["python3"], "env": ["VENICE_API_KEY"] }
export VENICE_API_KEY="vn_your_key_here"Promotion of Uncensored Content Generationmediumoversight evasion
The skill explicitly promotes 'uncensored' AI models, stating 'no content filters, no refusals.' This framing could encourage the generation of content that violates ethical guidelines, produces harmful outputs, or assists in malicious activities, potentially bypassing human oversight and safety mechanisms.
90% confidenceline 33
Venice is also **uncensored**: no content filters, no refusals, no "I can't help with that." Uncensored responses — No guardrails blocking legitimate use cases. --no-venice-system-prompt
Description Mismatchinfodescription mismatch
The skill's stated purpose heavily emphasizes 'true privacy' and 'zero data retention'. While the skill provides options to enable Venice AI's claimed privacy features (e.g., E2EE, private models), the skill itself is a client that sends user data (prompts, files) to the remote Venice AI service. The privacy guarantees are entirely dependent on the Venice AI service's implementation and trustworthiness, not on the local execution of the skill.
85% confidence
The skill's scripts make API calls to `venice.ai` and pass user data. The privacy claims are attributes of the remote service, not the local client skill.
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/clawhub/jonisjongithub/venice-ai)HTML
<a href="https://mondoo.com/ai-agent-security/skills/clawhub/jonisjongithub/venice-ai"><img src="https://mondoo.com/ai-agent-security/api/badge/clawhub/jonisjongithub/venice-ai.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/clawhub/jonisjongithub/venice-ai.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow