Definition
llms.txt and agents.md
Files on your site that tell AI agents who you are and how to interact with your catalog. Shopify now ships defaults for both.
From the episode “$7,500 AI Team Replacing an Entire Growth Department”Hear it on the podcast1 episode
Learn the language of AI-native founders
Every week: the new terms, tools, and workflows founders are using, straight from the podcast.
Related terms in Agents & Automation
In-House AgentKindling Solutions's smaller entry point: one AI role installed to solve one defined problem.Knowledge Base TrainingThe process of feeding audio recordings, sales scripts, and objection handling examples into an AI agent so it adopts the cadence, tone, and approach of a specific human salesperson.Knowledge Worker AutomationThe application of AI to tasks historically performed by white-collar professionals who work with information rather than physical materials - underwriters, analysts, lawyers, accountants, researchers, estimators. Unlike factory automation (which replaces physical labor) or earlier software automation (which replaced rule-based clerical tasks), knowledge worker automation targets judgment-adjacent tasks: reading documents, extracting information, summarizing findings, routing decisions, and populating systems. Mehul Shah's framing: the goal is not to replace the knowledge worker's judgment but to remove the tedious, error-prone, high-volume execution work from their critical path - freeing them to apply their expertise to the decisions that actually require it.Lifecycle Agentic FlowEhsan Mirdamadi's term for AI tools that span the entire software development lifecycle: writing code, committing to repos, running code reviews, debugging, and iterating. Distinguished from point solutions that assist at one step. Devin is the primary example. Designed for engineering teams who already have architectural context and a development workflow; not designed for non-technical founders who need requirements generation and full-stack scaffolding.Long-horizon agentsAgents that run continuously over extended periods rather than completing single tasks.Machine-readable inventoryA continuously updated agent list that software can query.MCP (Model Context Protocol)A standard that lets AI assistants connect to tools and data sources.MCP connectorIntegration layer that pipes CRM data such as Salesforce or HubSpot into an LLM.