Skip to content

Agentic Software Engineering Series

SASE uses agentic software engineering to mean software work where AI agents operate inside durable engineering systems: plans, work queues, review records, tests, commits, dependencies, and handoffs. The series explains why that coordination layer matters and how SASE implements it.

The canonical posts live on the SASE Blog. This page is the series hub for readers who want the launch arc in order.

Series Track

Order Post Status
1 Why Coding Agents Need Orchestration Launch stub
2 ChangeSpecs: Turning Agent Output Into Reviewable Work Planned
3 Beads: Dependency-Aware Work Units For Multi-Agent Execution Planned
4 XPrompts: Reusable Workflows Above One-Off Prompts Planned
5 ACE And AXE: Operating Agents From A Durable Control Plane Planned

Reader Paths

Start with the launch essay for the motivation, then move to the product surfaces that make the ideas concrete:

  • ACE TUI for the interactive control surface.
  • Spec-Driven Development for plans, epics, and executable phase work.
  • Beads for issue-like work items and dependency ordering.
  • XPrompts for reusable prompt and workflow packaging.
  • GitHub for source, issues, and implementation details.

Publishing Notes

The blog is the canonical publishing surface for SASE essays. Series posts should keep stable slugs, include frontmatter dates and categories, and link back here so readers can continue through the launch track without relying on archive pages alone.