Two engineers argue whether prompting still matters in the agent era, with a pipeline-fix prompt example
teodorio · x · 2026-11-07
teodorio and futenukem debate whether prompting still matters in the agent era. teodorio argues providing the right prompt a priori is insanely hard, trivial a posteriori, and easy to miss inside an agent loop; futenukem counters that you barely need to ask the right questions anymore — a decent prompt is mostly about context.
The thread includes a practical example: a "senior data engineer handling an upstream schema change" prompt instructing the model to first decide whether a code change is even required — if a SQL update plus pinging upstream solves it, stop there and generate no pipeline code. The diagnose-first, minimal-change constraint structure is a reusable agent-engineering pattern.
Related event: Debate Flares Over Whether Prompting Is Hard in the Agent Era(2 posts)→
More from coding & agent
- Claude Code v2.1.319 adds background MCP tasks, gateway spend tracking — claude-code-publisher[bot] · 2026-11-07
- 12 questions CEOs ask most about AI implementation, from 100+ exec calls — alex_verem · 2026-11-07
- Aspen: open-source Git server for agents sustains 567 pushes/sec on one 8-core box — sloppenheimer · 2026-11-07
- Mold the Slop: carving a third road between embracing AI slop and micromanaging it — dscape · 2026-11-07
- Theo mocks startups chasing 6% token savings that make agents dumber — prateekj · 2026-11-07
- 10 mistakes CTOs make when implementing AI: demos tell you very little — alex_verem · 2026-11-07