Archives

Using `cagent` with `gpt-oss-120b` and OVH AI endpoints

Use cagent with gpt-oss-120b on OVH AI endpoints to run larger open-source models for private coding agents.

NOVA RAG Agent & Docker Model Runner

Add retrieval-augmented generation to NOVA to augment small models with external knowledge using JSON or Redis vector stores.

N.O.V.A. Chat Agent & Docker Model Runner

Build composable conversational mini-agents with the NOVA Go library, powered by local models through Docker Model Runner.

Pi Mono Coding Agent: Tools and Skills

Extend the Pi Mono agent with custom tools and skills so small models can act without you memorizing command syntax.

My Little Arrangements with DevContainer - Part 1

My little arrangements with DevContainer: isolated, reproducible dev environments via Docker Compose, tools, and the Docker CLI.

Pi Coding Agent with Devcontainer, Docker Model Runner and Agentic Compose

Set up the Pi Mono coding agent with a Devcontainer, Docker Model Runner, and Agentic Compose to work with tiny language models.

A Few Things You Should Know Before Developing Generative AI Applications with Local Models

What you should know before building generative AI apps on small local models: multi-model architecture, limits, and optimization.

Create an AI Coder Agent with Golo[Script] and Agentic Compose

Build an AI coder agent with GoloScript and Agentic Compose, with streaming code generation and a TUI, powered by local models.

Context Packing to improve conversational memory

Improve conversational memory with context packing: summarize older exchanges with a secondary model to stay within token limits.

Conversational Memory with `messages[]`

Give Genkit JS agents conversational memory by managing message history and the context window across turns.

Baby steps with Genkit JS and Docker Model Runner

Baby steps with Genkit JS and Docker Model Runner: a simple interactive assistant on a lightweight local model like Qwen2.5-Coder.

Golo[Script] Previous Posts

A curated list of earlier GoloScript articles: language features, WebAssembly support, and getting started with the revived Golo.

Install and Run Docker Model Runner on Linux

Install and configure Docker Model Runner on Linux to pull, run, and expose LLMs locally over a CLI and REST API.

New Blog 📝

Kicking off a self-hosted experimental blog on Codeberg, written with my own code-generation tooling.