Engineering Notes

Engineering Notes

Thoughts and Ideas on AI by Muthukrishnan
15 Mar 2026

Working Memory Compression and Context Distillation in Long Horizon Agents

How long-running agents compress, distill, and selectively retain working memory to operate effectively within finite context windows
02 Mar 2026

Spec-Driven Development with spec-kit: Stop Vibe Coding, Start Specifying

A complete tutorial on using GitHub's spec-kit to bring structure to AI-assisted development — from install to your first specification.
28 Feb 2026

Conceptual AI Agents Universe — System Design Document

A plugin-based platform architecture where each AI agent system is an independently subscribable capability. One interface, one orchestrator, unlimited agents — added without touching existing code.
19 Feb 2026

Agent Evaluation and Benchmarking for Measuring What Matters

Learn how to systematically evaluate AI agent performance using benchmarks, metrics, and evaluation frameworks that go beyond simple accuracy
17 Feb 2026

What If GitHub Was Built for AI Agents

Reimagining source code management from the ground up for AI agents, with intent based commits, simulation before merge, agent reputation, and automatic rollback contracts
15 Feb 2026

Knowledge Distillation Teaching Smaller Agents From Larger Ones

Learn how knowledge distillation enables large, expensive AI agents to teach smaller, faster ones — reducing cost and latency while preserving capability
02 Feb 2026

Claude Code CLI Best Practices Checklist

A comprehensive guide to tips, tricks, and best practices for maximizing productivity with Claude Code. --- ## Table ...
02 Feb 2026

The Complete Guide to Git Worktrees with Claude Code

## Table of Contents 1. [Introduction](#1-introduction) 2. [Understanding Git Worktrees](#2-understanding-git-worktree...
01 Jan 2026

Building a Multi-Agent Swarm with OpenAI Swarm Framework

Learn the basics of OpenAI's experimental Swarm framework for lightweight multi-agent orchestration with handoffs and routines.
31 Dec 2025

Adding Context to Chunks for Better Retrieval with Contextual RAG

Improve your RAG pipeline by prepending contextual information to each chunk before embedding, reducing retrieval failures by up to 67%.