by guaardvark · MCP Server · ★ 264
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is guaardvark safe to install? View the security audit →
Guaardvark See the VERSION file for the current release · guaardvark.com The self-hosted, offline-first AI workstation. Autonomous agents that see a real virtual desktop and control apps. A three-tier neural router (AgentBrain). Parallel coding agent swarms in isolated git worktrees. Local video (Wan 2.2, CogVideoX), 4K/8K upscaling, full-song music + neural voice, RAG over your documents, voice chat, and a 70+ tool engine — everything on your hardware. Your machine. Your data. Your rules. For the exhaustive feature list, models, surfaces, and plugin details, see CAPABILITIES.md. This README focuses on the marquee experience, quick start, and what makes Guaardvark different. Highlights (as of latest release) Video & Audio Production — Wan 2.2 (T2V + I2V), CogVideoX, SVD; ACE-Step full-song generation with LLM tag polish; Chatterbox/Kokoro neural voice + Piper; explicit consent-gated voice cloning; frame-by-frame 4K/8K upscaling. AgentBrain + Screen Agents — Reflex/Instinct/Deliberation routing. Agents drive a real Ubuntu/XFCE desktop on a virtual display (), see with vision models (Gemma4 native ), use closed-loop servo targeting, and stream per-step reasoning.
git clone https://github.com/guaardvark/guaardvark.git && cd guaardvark && ./start.sh
git clone https://github.com/guaardvark/guaardvark.gitTaken from the project's README on GitHub. Check the security audit before running it.
No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →
| Stars | 264 |
| Forks | 55 |
| Language | Python |
| Category | MCP Server |
| License | MIT |
| Quality Score | 64.8691400787261/100 |
| Open Issues | 12 |
| Last Updated | 2026-10-15 |
| Created | 2026-03-11 |
| Platforms | browser, claude-code, mcp, python |
| Est. Tokens | ~22k |
Looking for a guaardvark alternative? If you're comparing guaardvark with other mcp server tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
Local-first, agent-native control plane for ComfyUI — MCP server + sidebar agent that generates images, video
Open‑WebUI Tools is a modular toolkit designed to extend and enrich your Open WebUI instance, turning it into
c4 GenAI Suite
Session-scoped Claude Code agent for ComfyUI video production — 12 specialized skills for image gen, video, vo
The whole local AI stack in one executable: it runs and manages local AI models across every GPU, and it's a s
Give Claude Code, Cursor, Codex CLI a ChatGPT for your codebase. Multi-agent knowledge engine, grounded Q&A wi
Explore other popular mcp server tools:
guaardvark is Self-hosted AI studio on your own GPU: chat with your files (RAG), screen and browser agents, coding swarms, LoRA training, an MCP server for Claude Code and Cursor, and local video, image, music and . It is categorized as a MCP Server with 264 GitHub stars.
guaardvark is primarily written in Python. It covers topics such as ai-agents, claude-code, claude-code-plugin.
The README on GitHub (github.com/guaardvark/guaardvark) gives: git clone https://github.com/guaardvark/guaardvark.git && cd guaardvark && ./start.sh , then git clone https://github.com/guaardvark/guaardvark.git. Usage details are in the same README.
guaardvark is released under the MIT license, making it free to use and modify according to the license terms.
The top alternatives to guaardvark on Agent Skills Hub include comfyui-mcp, open-webui-tools, c4-genai-suite. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: