Measuring LLM Capability and Reliability: Pass@k vs Pass^k
The truth about working with LLMs
Shaswat Rungta
Personal blog about my experiments with tools and frameworks.
Latest posts
When 💩 hits the database
Chapter 1: A brief history of world knowledge
Running Multiple App Copies Locally with Git Worktrees and Caddy
📚 Companion post: Git Worktrees for Agentic Development explains why worktrees are useful when people or coding agents work on several branches at once.
Git Worktrees for Agentic Development
The problem agents create
My Cache Has Trust Issues - A Therapist’s Guide to Cache Invalidation
Session 1: Meeting Your Cache
My Cache Has Trust Issues - A Therapist’s Guide to Cache Invalidation
Session 1: Meeting Your Cache
The Internet is Held Together by Duct Tape
The tech debt of the Internet
The Internet is Held Together by Duct Tape
The tech debt of the Internet
Making Your Codebase LLM Friendly
What Does “LLM Friendly” Really Mean?
MCP server basics - Building an MCP server for reminders
Introduction to MCP Servers