AI-native SDLC: four agent-driven stages, four human gates between code and production
Pavan_Belagatti · x · 2026-11-17
Pavan Belagatti argues that while code is becoming cheap, shipping reliable software remains hard — the bottleneck is waiting: tickets, reviews, and handoffs that add only delay.
- The four stages stay the same — Plan, Build, Ship, Operate — with an agent doing the heavy lifting in each: a spec agent drafts requirements, a coding agent writes and tests code, a release agent prepares deploys, an SRE agent watches production
- The accompanying diagram lays out the full pipeline: each stage has a Human Gate and an artifact (spec.md, PR, Release, Insights), and Operate insights feed back into Planning
- Core claim: skipping human gates is a disaster — developers approve the spec, the PR, and the release, and decide which fixes matter most
A methodology piece on running agent-driven development while keeping humans at the control points.
More from coding & agent
- LangChain hosts SF hands-on workshop on Managed Deep Agents with LangSmith — LangChain · 2026-11-18
- Vite+ 1.7.0 adds type-aware linting via tsrs, a Rust port of TypeScript 7 — cnakazawa · 2026-11-18
- Higgsfield's new AI film: 'The hardest part is making it not look like AI' — CurieuxExplorer · 2026-11-18
- Project triangle flips in the AI agent era: defining 'good' becomes the job — lak_luster · 2026-11-18
- The levels of agentic engineering: find what you still do by hand, then automate it — rseroter · 2026-11-18
- GitHub Next's Maggie Appleton on speculative design for prototyping software futures — JayAlammar · 2026-11-18