Storage Management

Stackie tracks initialization state for each block’s storage so that re-running stackie up skips setup steps that have already completed (e.g., database schema creation, seed data loading).

Storage Identity

Each block’s storage is identified by {project}/{block}:

Stack block: project is the stack name (e.g., supabase/postgres)
Standalone block: project is standalone (e.g., standalone/postgres)

Directory Structure

~/.stackie/storage/
├── volumes/{project}/{volume_name}/    # Shared project volumes
│   └── _stackie_init.json             # Init state metadata
└── paths/{sandbox_name}/               # Per-sandbox isolated paths
    └── _stackie_init.json             # Init state metadata

Cleaning Storage

To force re-initialization of a block’s storage:

# Remove stored data for a specific block
stackie storage clean supabase/postgres

# Remove all storage for a project
stackie storage clean --project supabase

# Remove only archived runs (from --fresh)
stackie storage clean --archived