proj
AI Agent Evals
Security-focused evaluation benchmarks for AI coding agents. Tests whether agents can perform real-world incident response — investigating a compromised production server, removing attacker backdoors, restoring SSL/TLS chains, fixing nginx configs, and hardening Go services — all inside Docker-based infrastructure using the Harbor evaluation framework.
- AI Agents
- · Security
- · Evals
- · Python
- · Docker
- · Go
proj
InfAI
Zero-management launch templates for llama.cpp. Auto-detects GGUF models, saves named profiles in SQLite, and launches llama-server with live logs and real-time metrics — no flag soup. Supports 11+ themes on macOS and Linux.
- Go
- · TUI
- · llama.cpp
- · LLM
- · SQLite
proj
AI Agents Skill Library
A curated skill library for AI agents, enabling modular capability composition and reuse across agent architectures.
proj
Chat Bot using GPT-2
HTTP server using Rust with a text-to-text generator bot via HuggingFace. Multiple implementations: WasmEdge, Spin by Fermyon, and Tokio/Async.
- Rust
- · WASM
- · GPT-2
- · HuggingFace