Vadym Parakonnyi voznik

LLM Wiki v2

A pattern for building personal knowledge bases using LLMs. Extended with lessons from building agentmemory, a persistent memory engine for AI coding agents.

This builds on Andrej Karpathy's original LLM Wiki idea file. Everything in the original still applies. This document adds what we learned running the pattern in production: what breaks at scale, what's missing, and what separates a wiki that stays useful from one that rots.

What the original gets right

The core insight is correct: stop re-deriving, start compiling. RAG retrieves and forgets. A wiki accumulates and compounds. The three-layer architecture (raw sources, wiki, schema) works. The operations (ingest, query, lint) cover the basics. If you haven't read the original, start there.

Running OpenCode with Gemma 4 26B on macOS (via llama.cpp)

As of April 2026, Gemma 4 tool calling is broken in Ollama v0.20.0 (ollama/ollama#15241) - the tool call parser fails and streaming drops tool calls entirely. OpenCode also has issues with local OpenAI-compatible providers (anomalyco/opencode#20669, #20719).

This guide documents a working setup using:

llama.cpp (built from source with PR #21326 template fix + PR #21343 tokenizer fix) instead of Ollama
OpenCode built from source with PR #16531 tool-call compatibility layer

Tested on macOS Apple Silicon (M1 Max, 32GB) on April 2, 2026.

Verified Spec-Driven Development (VSDD)

The Fusion: VDD × TDD × SDD for AI-Native Engineering

Overview

Verified Spec-Driven Development (VSDD) is a unified software engineering methodology that fuses three proven paradigms into a single AI-orchestrated pipeline:

Spec-Driven Development (SDD): Define the contract before writing a single line of implementation. Specs are the source of truth.
Test-Driven Development (TDD): Tests are written before code. Red → Green → Refactor. No code exists without a failing test that demanded it.

Knowledge Hub: Technical Deep Dive

A companion to "Building a Knowledge Hub in 13 Days"

What It Is

Knowledge Hub is an enterprise knowledge aggregation platform that syncs data from multiple sources into a searchable vector database with AI-powered retrieval-augmented generation (RAG).

Claude Agent SDK Technical Specification

Version: 1.0.0 Protocol Version: 2024-11-05 Last Updated: 2026-01-10

Agent Guidelines for Rust Code Quality

This document provides guidelines for maintaining high-quality Rust code. These rules MUST be followed by all AI coding agents and contributors.

Your Core Principles

All code you write MUST be fully optimized.

"Fully optimized" includes:

Inferal Workspace Architecture: How We Work at Inferal

(Blog version available at https://inferal.com/blog/workspace-architecture/)

Update: Inferal Workspace is going open source!

Your org's brain that AI can use

This is not our core product. This document describes our internal operating environment - how we run the company. We share it to show the environment you'd join and demonstrate our philosophy in action. For what we're building, see What We're Building below.

Here's my AGENTS.md (also linked from CLAUDE.md as @AGENTS.md) for hacking agentically on MDFlow recipes.

I have this in ~/.mdflow/, and the agents/recipes live in ~/.mdflow/agents/ and added to the path so that they can be invoked as commands.

With this I can use a coding agent like Claude Code or GitHub Copilot in VSCode and say something like:

> create a new agent using copilot that reviews all the code files in this directory as a poem

	# Create a new worktree and branch from within current git directory.
	ga() {
	if [[ -z "$1" ]]; then
	echo "Usage: ga [branch name]"
	exit 1
	fi

	local branch="$1"
	local base="$(basename "$PWD")"
	local path="../${base}--${branch}"

	{{ states.binary_sensor
	\| selectattr('attributes.device_class', 'in', ['door','window'])
	\| selectattr('state', 'equalto', 'on')
	\| list \| count }}