collection-sync

name	collection-sync
description	Keeps the Allrecipes aggregator in sync with its 3 source repositories (Grandmasrecipes, Grannysrecipes, MomsRecipes). Detects changes, identifies new/modified/deleted recipes, validates cross-collection indexes, and checks for duplicates.
version	1.0.0

Collection Sync

Three kitchens, one family cookbook.

Purpose

Allrecipes is the aggregator hub for three family recipe collections. This skill ensures the aggregator stays in sync with the source repos, catches duplicates across collections, and maintains cross-collection search indexes.

When to Fire

• On /sync command
• At session start (quick check)
• After bulk recipe changes in any source repo
• When cross-collection features are being worked on

Source Repositories

Collection	Repository	Owner
Grandma Baker	/home/user/Grandmasrecipes/	Grandma Baker's recipes
Granny Hudson	/home/user/Grannysrecipes/	Granny Hudson's recipes
MomMom Baker	/home/user/MomsRecipes/	MomMom Baker's recipes

Available Scripts

# Aggregate all collections
python3 scripts/aggregate_collections.py

# Aggregate cooking tips
python3 scripts/aggregate_tips.py

# Find duplicates across collections
python3 scripts/analyze_duplicates.py
python3 scripts/check-duplicates.py

# Build ingredient index
python3 scripts/build-ingredient-index.py

# Build search shards
python3 scripts/build_shards.py

Sync Workflow

1. Quick Status Check (Session Start)

For each source repo, compare:

• Last commit timestamp in source repo vs. last sync timestamp
• Recipe count in source vs. count in aggregator
• Any new files in source data/ directories

Report:

## Sync Status — [date]
| Collection | Source Recipes | Aggregated | Last Sync | Status |
|-----------|--------------|-----------|-----------|--------|
| Grandma Baker | 245 | 245 | 2026-03-20 | ✅ Current |
| Granny Hudson | 180 | 178 | 2026-03-15 | ⚠️ 2 new recipes |
| MomMom Baker | 150 | 150 | 2026-03-22 | ✅ Current |

2. Full Sync

# Step 1: Check for changes
python3 scripts/check-duplicates.py  # Find cross-collection duplicates first

# Step 2: Aggregate
python3 scripts/aggregate_collections.py

# Step 3: Rebuild indexes
python3 scripts/build-ingredient-index.py
python3 scripts/build_shards.py

# Step 4: Aggregate tips
python3 scripts/aggregate_tips.py

3. Duplicate Detection

Cross-collection duplicates are expected — the same dish may appear in multiple grandmothers' collections. Handle them as:

• Same dish, different recipe — Keep both, note the cross-reference ("See also Granny Hudson's version")
• Identical recipe — Flag for review (which collection should be primary?)
• Similar name, different dish — Clarify in metadata

python3 scripts/analyze_duplicates.py

4. Validation

After sync, verify:

• All source recipes are represented in the aggregator
• No orphaned entries (recipe in aggregator but deleted from source)
• Cross-collection search returns results from all 3 collections
• Ingredient index includes all unique ingredients
• Search shards are current

Sync Status Tracking

Maintain a sync status file or memory entry:

{
  "last_sync": "2026-03-25T14:30:00Z",
  "collections": {
    "grandma": {"recipe_count": 245, "last_commit": "abc123"},
    "granny": {"recipe_count": 180, "last_commit": "def456"},
    "mommom": {"recipe_count": 150, "last_commit": "ghi789"}
  }
}

Integration

• cognitive-memory — Remember sync state across sessions
• recipe-validation — Validate aggregated data after sync
• nutrition-estimator — Ensure nutrition data propagates to aggregator
• careful-not-clever — Never merge recipes without checking. Duplicates need human review.

What NOT to Do

• Never delete a recipe from a source repo through the aggregator
• Never modify source recipes from the aggregator (read-only from sources)
• Never assume two similarly-named recipes are the same without checking
• Never run a full re-aggregate without checking duplicates first

---

Soli Deo Gloria — Three grandmothers' legacies, faithfully preserved together.