Agent infra builder's 3-question test: when a plain script beats an agent
uriwa · reddit · 2026-09-03
A developer who builds agent infra shares a 3-question framework for choosing between an agent and a plain script:
- Is the procedure known? If you can write the steps ahead of time, a script is faster, cheaper and deterministic. Agents shine when steps must be figured out reactively.
- How many items? Agents suit one complex case; at 10k items, per-item LLM latency and token costs add up while a script finishes in seconds.
- Are items independent? Mixing unrelated items in one context leaks details and conflates customers/errors; code handles independent items trivially.
Sweet spot: unknown procedure, small scale, interrelated items. He also notes the cost angle: you're paying per token to reason about things that may not need reasoning, and seconds-vs-milliseconds matters to a waiting user.
More from coding & agent
- Two Attribution Disputes Later: Keeping Open Source Research Viable in the Age of Coding Agents — dhruv2038 · 2026-09-03
- Meta's Muse Spark 1.3 is now free on OpenCode — ramagetime · 2026-09-03
- GitHub Announces First-Ever Copilot Day on September 10 With Product Announcements — PaulShellDev · 2026-09-03
- Inside the $950 Maven Course on Production-Ready Enterprise AI Agents — hugobowne · 2026-09-03
- Deploying an AI Agent on Modal in Six Minutes: A Walkthrough — hugobowne · 2026-09-03
- Dev Offers Scale AI 300GB of Agent Session Logs, Runs 63 Agent Subscriptions at $13.5K/Month — doodlestein · 2026-09-03