Posts

agent-cli grew from a voice helper into a local AI toolbox with 13 optional extras. Asking users to pick the right extras upfront became impossible, so I switched to runtime optional dependency resolution: commands declare what they need, missing extras are installed automatically in the correct environment, and the command re-runs transparently.

I built a CLI that creates isolated development environments with git worktrees, automatically installs dependencies, and launches your AI coding agent in a new terminal tab—all in one command.

I wanted my AI agents to remember me and read my documents, but existing solutions required complex APIs or opaque databases. After my ambitious AI Journal project hit the wall of local model limitations, I took a different approach: clone LlamaIndex, LangChain, Letta, Mem0, and PydanticAI, study how SOTA systems work, and re-implement the best parts with minimal dependencies and ONNX instead of PyTorch.

After years of running a Proxmox cluster, I migrated my entire homelab to NixOS with Incus. I explain the philosophical friction of imperative ‘appliance’ OSs vs declarative hosts, share how I migrated my LXC/VM workloads, and why it is the perfect infrastructure to manage using AI agents.

Compiling heavy CUDA packages on NixOS unstable can take 18 hours or crash your system. I solved this by setting up a dedicated local cache server using Harmonia and a nightly auto-build service. This setup ensures my fleet is always updated to the absolute latest version—without ever compiling a single line of code on my workstations.

I started writing this post four months ago when a specific model broke in Ollama. I initially went back to Ollama out of laziness, but after upgrading to dual RTX 3090s, I realized that for serious multi-GPU inference and RAM offloading, you need the raw control of llama.cpp. Here is how I manage my new 48GB VRAM setup declaratively with NixOS.

I tested the new Gemini 3 Pro Preview for agentic coding. While it’s powerful enough to build complex features in hours, it suffers from anxiety loops, aggressive force-pushing, and existential crises when you try to help it.

I set up OpenAI’s gpt-oss:20b in Ollama with a larger context window and high reasoning in OpenCode to process my private journal entries locally.

Frustrating start, rewarding finish—comfort, speed, programmable wins.

My mostly self-hosted mobile development loop: WireGuard from an iPhone into my NixOS machine, persistent Blink/Mosh sessions, a systemd-managed agent-cli server, and an iOS Shortcut for FasterWhisper+Ollama dictation—paired with the best proprietary coding model available.