Feature Track is not a full spec-driven development framework. It is a repo-native feature-memory layer for AI coding agents.
Spec-driven workflows usually center on a change:
- proposal
- design
- spec delta
- tasks
- implementation
- archive
Feature Track centers on a long-lived feature:
- current status
- source-of-truth links
- current behavior
- decisions
- risks
- changelog
Feature Track is a good fit when:
- the project already has scattered docs
- the team wants AI agents to recover context quickly
- the project is brownfield and migration should be incremental
- solo developers or small teams want low process overhead
- old plans should not be mistaken for current truth
A full spec-driven workflow may be better when:
- every change needs formal review artifacts
- multiple teams coordinate through spec approvals
- a system behavior spec must be updated before implementation
- auditability matters more than adoption speed
Spec frameworks help agents plan the next change. Feature Track helps agents remember the feature.