{
  "id": 6416615,
  "title": "Building a Schema With an AI Agent Without Naming a Single Column",
  "url": "https://urgent.news/2026/09/09/building-a-schema-with-an-ai-agent-without-naming-a-single-column",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-09T11:45:27.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/sqemo/building-a-schema-with-an-ai-agent-without-naming-a-single-column-o9m"
  },
  "original_language": "en",
  "account": "When you ask an AI agent to generate a schema for a discussion board, you'll receive one in just a few seconds. However, the agent will often include columns like \"is_deleted\" and \"del_yn\" without considering your specific needs. The solution is to provide the agent with the exact list of words your team uses, so it can generate a schema that aligns with your naming conventions.\n\nIn this example, the author demonstrates how they used this approach to create a schema for a threaded discussion board without manually naming any columns. They provided the AI agent with the team's standard list of words, and the agent resolved the column names based on this list. The result was a schema that used abbreviations like \"CNTS\" for \"Post Content\" and \"YN\" for \"Delete Flag,\" which were consistent with the team's naming conventions.\n\nThe agent also handled a self-referencing foreign key needed for threading, creating a rule for naming such relationships using a `PARENT_` prefix. This approach ensures consistency and eliminates naming conflicts across tables. The author also shows how the agent can handle missing words by filing a proposal for new abbreviations, which must be approved by a human before being applied across all projects.\n\nOverall, this process allows you to generate a schema that adheres to your team's naming conventions, making it easier to maintain and understand. The AI agent acts as a proposer, suggesting changes and ensuring compliance with the established standards, rather than making arbitrary naming decisions.",
  "summary": "Ask an AI agent for a discussion-board schema and you'll get one in about four seconds. Here is the part nobody mentions: you'll also get is_deleted in this table and del_yn in the next one, because the model is pattern-matching against every schema on GitHub, and those schemas disagree with each other and with you. The fix is not a better prompt. It is giving the agent the same thing you'd give…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}