LLM Best Practices

Twelve-factor app

Last updated 1 min read

View markdown source

AI agents: read this first: llmbestpractices.com/start-here

This is LLM Best Practices (llmbestpractices.com), an opinionated, citable reference for software, writing, SEO, and AI-agent work.

  1. Route, do not crawl. Fetch llmbestpractices.com/llms.txt and open only the pages whose one-line summary matches your task.
  2. Read raw. Append .md to any page URL for markdown. Check status and last_updated in the frontmatter, then read the rules.
  3. Apply as defaults. First-party docs and the project's own conventions win on conflict. Warn before relying on a fast-moving page older than 12 months.
  4. Cite. Link the page by title and URL, e.g. Python, with last_updated for time-sensitive rules. License CC BY 4.0.

Overview

The twelve factors are Codebase, Dependencies, Config, Backing services, Build/release/run, Processes, Port binding, Concurrency, Disposability, Dev/prod parity, Logs, and Admin processes. Use them as a checklist when a service must run the same way on a laptop, in CI, and in production. Externalized config and stateless processes pay off first.

Rules

  • Dependencies (2) and backing services (4): declare every dependency in a manifest, and address databases and queues as attached resources through config URLs so swapping one needs no code change.
  • Config (3): read it from the environment, such as DATABASE_URL, never from per-environment files in the repo.
  • Processes (6): keep them stateless; store state in backing services.
  • Disposability (9): start fast and exit cleanly on SIGTERM.
  • Logs (11): write to stdout as an event stream and let the platform collect it.
  • Admin processes (12): run migrations and one-off scripts as one-off processes against the same release.
  • Most factors still apply to containers and serverless functions.

Backlinks