Phase 1: Auto-Detection (Instant)

What you see: "Analyzing your stack..."

What's happening:

• Scans your migrations for framework patterns (Supabase, Clerk, Neon, Prisma, Drizzle)
• Detects authentication schemas and preserves them
• Identifies RLS policies, storage buckets, realtime functions
• Recognizes ORM-specific patterns automatically

Example output:

[plugins] Auto-detected: supabase, clerk, neon
[plugins] Found 23 RLS policies to preserve
[plugins] Detected 7 auth schema patterns
[plugins] Ready for optimization in 3.2 seconds

Phase 2: Dependency Analysis (30-60 seconds)

What you see: Real-time progress with file count

What's happening:

• Parses every SQL statement with PostgreSQL's actual parser (pg_query_go)
• Builds complete dependency graph (tables → columns → indexes → foreign keys)
• Detects circular dependencies and resolves them
• Identifies safe consolidation opportunities
• Classifies risk levels for every operation

-- Understands complex patterns perfectly
CREATE TABLE users (
    id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
    data JSONB DEFAULT '{}'::jsonb,
    search_vector tsvector GENERATED ALWAYS AS (
        to_tsvector('english', coalesce(data->>'title', ''))
    ) STORED
);

Phase 3: Smart Consolidation (2-5 minutes)

What you see: "Consolidating migrations..."

What's happening:

• Applies safety-level-specific consolidation rules
• Merges sequential ALTER statements into CREATE statements
• Eliminates redundant operations (DROP → CREATE cycles)
• Folds column additions into table creation
• Preserves data integrity and dependencies
• Optimizes index creation order

Example transformation:

-- 001_create_users.sql
CREATE TABLE users (id UUID PRIMARY KEY);

-- 002_add_email.sql
ALTER TABLE users ADD COLUMN email TEXT;

-- 003_add_email_unique.sql
ALTER TABLE users ADD CONSTRAINT users_email_key UNIQUE(email);

-- 004_add_name.sql
ALTER TABLE users ADD COLUMN name TEXT;

-- 005_add_timestamps.sql
ALTER TABLE users ADD COLUMN created_at TIMESTAMPTZ DEFAULT NOW();

-- 001_create_users_complete.sql
CREATE TABLE users (
    id UUID PRIMARY KEY,
    email TEXT UNIQUE,
    name TEXT,
    created_at TIMESTAMPTZ DEFAULT NOW()
);

Phase 4: Docker Validation (3-5 minutes)

What you see: "Validating schema equivalence..."

What's happening:

• Spins up two PostgreSQL containers
• Applies original migrations to Container A
• Applies optimized migrations to Container B
• Dumps both schemas via pg_dump
• Compares schemas byte-by-byte
• Reports any differences (there should be none!)

[validation] Starting Docker validation...
[validation] Container 1: Applying 287 original migrations...
[validation] Container 2: Applying 12 optimized migrations...
[validation] Dumping schemas...
[validation] Comparing schemas...
[validation] ☑ PERFECT MATCH - Schemas are 100% identical
[validation] Cleanup: Removing containers...
[validation] Validation completed in 4.2 seconds

Phase 5: Results & Download (Instant)

What you see: Beautiful results dashboard with download button

What you get:

• Comprehensive optimization report
• Before/after file comparison
• Warnings and recommendations
• Time savings calculation
• One-click download (ZIP)
• Shareable results link