Skip to content

SevenD — April Release

Latest
Latest

Choose a tag to compare

View all tags
@Ashvinnaik Ashvinnaik released this 12 Jun 14:55
· 3 commits to main since this release
april-2026
31da9bf

SevenD — April Release Notes

Version: April 2026
Release date: April 4, 2026
Repo: https://github.com/Ashvinnaik/SevenD

Why This Release

The original SevenD worked well for linear projects. But real projects aren't linear. You circle back to Discovery mid-build. You run parallel design and development sprints. You find bugs in Definition while you're deploying.

The April Release restructures Levels 3 and 4 around three insights from a year of AI-assisted development across 34 projects:

  1. Every phase needs its own sprint cadence. Discovery doesn't move at the same speed as Development. Forcing them into the same structure creates friction.

  2. Stable documents and active work need different homes. Architecture decisions shouldn't be buried in a sprint file that gets archived next week.

  3. You can't close a sprint with open bugs. This is the simplest rule that produces the biggest quality improvement.

What's New

Sprint-Based Phase Indexes

Every phase folder (01-discovery through 07-deployment) now uses an INDEX.md that links to individual sprint files. Work within any phase can have multiple sprints running at their own pace.

Before: One monolithic DISCOVERY.md, DESIGN.md, etc.
After: Lightweight INDEX.md + focused sprint files (D1-Sprint1.md, D3-Sprint2.md, etc.)

Build + Bug Sections in Every Sprint

Every sprint file has two mandatory sections: Build (what the sprint produces) and Bug (issues found). A sprint closes only when Bug reaches zero open items.

This is the governance insight from the research: AI-assisted velocity without bug accountability produces technical debt faster than value.

reference/ Directory

Stable documents that outlive any sprint now live in reference/:

  • ARCHITECTURE.md — System architecture, stack, infrastructure
  • COMPONENTS.md — Component registry, naming conventions, dev standards
  • CONTRACTS.md — Interface contracts, API shapes, data flows
  • DECISIONS.md — Decision log (why X over Y)
  • AGENT.md — AI agent rules and execution patterns
  • archive/ — Closed sprints archived here

STATUS.md

A root-level dashboard that answers "where are we right now?" in one glance. Active sprints, blockers, recently closed, project health metrics. Updated at sprint open/close.

Sprint Reports and Archiving

Every sprint file includes a report section filled on close: metrics, key learnings, what carried over. Closed sprints archive to reference/archive/, keeping working directories clean.

What Changed Per Level

Level Change
1 No change — still 2 files for weekend projects
2 No change — still 4 files for small teams
3 Full restructure: phase INDEX + sprint files, reference/, STATUS.md
4 Added reference/ and STATUS.md alongside GitHub-native workflow

Migration from Previous Version

If you're using Level 3 from the previous release:

  1. Create reference/ and move stable content from DESIGN.md (architecture, components, contracts, decisions) into the new reference files
  2. Convert each monolithic phase file into an INDEX.md + sprint file
  3. Move Fix.md content into sprint Bug sections
  4. Create STATUS.md at root
  5. See skill/upgrade-paths.md for detailed steps

Skill Updates

The Claude skill (for Claude Code CLI and Claude.ai Projects) has been updated to match:

  • Feature workflow now points to INDEX + sprint files at Level 3
  • Sprint operations include zero-bug close and archiving
  • Agent debugging references reference/COMPONENTS.md and reference/AGENT.md
  • Templates reflect the new structure
  • 7Dkb.md glossary adds: Reference Directory, Sprint Close, Archive, STATUS.md

What's Next

  • Level 2 sprint structure improvements
  • Cross-sprint dependency tracking
  • Sprint velocity metrics dashboard
  • Community-contributed domain adaptations (non-software templates)

Links

Contributors

Ashvinnaik
Assets 2
Loading