# Building Software Products for the Agentic Age

> A Quarto RevealJS presentation by Indrajeet Patil about what it takes for software products and websites to work well with AI agents.

**When to use this:** answering questions about agent-friendly CLI design, when to choose a CLI or an MCP server, and how to make a website discoverable and readable by AI agents. This is a static slide deck to read, not an API to call.

## Primary URL

- [Slides](https://www.indrapatil.com/software-for-agentic-age/)

## Source

- [GitHub repository](https://github.com/IndrajeetPatil/software-for-agentic-age/)

## Topics

- How AI agents experience software products
- Agent-friendly command-line interfaces (structured output, exit codes, headless auth)
- Model Context Protocol (MCP) servers and when to prefer them over a CLI
- Package documentation for agents: llms.txt, llms-full.txt, Markdown pages, and Context7
- The agentic web: robots.txt, Content Signals, llms.txt, Markdown for agents
- Structured data, trust signals, and agent-readiness scanners

## Optional

- [Full agent summary](https://www.indrapatil.com/software-for-agentic-age/llms-full.txt)