# Workplan 21: LadybugDB Backend Completion
> ⏸️ **STATUS: PENDING** (2025-12-08)
>
> **Priority**: 🟡 MEDIUM - Backend expansion
> **Prerequisites**: WP16 (SDK) and WP20 (Cloud) releases complete
> **Estimated Effort**: 14-21 hours
---
## Overview
Complete the LadybugDB backend implementation to production-ready status. The current skeleton implementation is missing critical GraphBackend methods and has security/quality issues identified in code review.
**Current State**: Skeleton with ~30% implementation
**Target State**: Full GraphBackend implementation matching FalkorDBLite quality
---
## Parallel Execution Guide
This workplan can be executed with **2 parallel agents** after Section 1 is complete.
### Dependency Graph
```
Section 1: Core Methods (SEQUENTIAL - must complete first)
│
├──► Section 2: Query/Search Methods ──┐
│ ├──► Section 4: Testing
└──► Section 3: Relationship Methods ──┘
│
▼
Section 5: Documentation
```
### Parallel Work Units
| Agent | Section | Dependencies | Can Run With |
|-------|---------|--------------|--------------|
| **Agent 1** | Section 1: Core Methods | None | Solo (prerequisite) |
| **Agent 2** | Section 2: Query/Search | Section 1 | Agent 3 |
| **Agent 3** | Section 3: Relationships | Section 1 | Agent 2 |
| **Agent 4** | Section 4: Testing | Sections 2-3 | Solo |
| **Agent 5** | Section 5: Documentation | Section 4 | Solo |
---
## Section 1: Core Infrastructure (CRITICAL)
### 1.1 Fix Parameterized Query Support
**File**: `/Users/gregorydickson/claude-code-memory/src/memorygraph/backends/ladybugdb_backend.py`
**Current Issue**: Parameters are ignored, creating SQL injection vulnerability.
**Tasks**:
- [ ] Research LadybugDB's parameterized query support
- [ ] If supported: Implement proper parameter passing in `execute_query()`
- [ ] If not supported: Implement query sanitization/escaping
- [ ] Add security warning to docstring if parameterization unavailable
- [ ] Update all query construction to use safe patterns
**Acceptance Criteria**:
- No string interpolation with user data in queries
- All queries use either parameterization or sanitization
- Security documented in module docstring
### 1.2 Fix Result Format Handling
**File**: `/Users/gregorydickson/claude-code-memory/src/memorygraph/backends/ladybugdb_backend.py`
**Current Issue**: Code expects dict results, but LadybugDB returns arrays.
**Tasks**:
- [ ] Verify actual result format from LadybugDB `get_next()` method
- [ ] Create `_result_to_dict(row, columns)` helper method
- [ ] Update `execute_query()` to return consistent dict format
- [ ] Fix `health_check()` result access (line 200)
- [ ] Add column name extraction from query results
**Reference**: See FalkorDBLite's result handling pattern
### 1.3 Add Type Conversion Helpers
**File**: `/Users/gregorydickson/claude-code-memory/src/memorygraph/backends/ladybugdb_backend.py`
**Tasks**:
- [ ] Implement `_ladybugdb_to_memory(data: dict) -> Memory`
- [ ] Implement `_memory_to_properties(memory: Memory) -> dict`
- [ ] Implement `_parse_datetime(dt_value) -> datetime`
- [ ] Implement `_ensure_utc(dt: datetime) -> datetime`
- [ ] Implement `_ladybugdb_to_relationship(data: dict) -> Relationship`
**Reference**: Copy patterns from `falkordblite_backend.py` lines 110-180
### 1.4 Fix Exception Types
**File**: `/Users/gregorydickson/claude-code-memory/src/memorygraph/backends/ladybugdb_backend.py:149-151`
**Tasks**:
- [ ] Change query execution errors from `SchemaError` to `DatabaseConnectionError`
- [ ] Reserve `SchemaError` for schema initialization failures only
- [ ] Review all exception usage for consistency
---
## Section 2: Memory CRUD Methods
### 2.1 Implement store_memory()
**Tasks**:
- [ ] Implement `async def store_memory(self, memory: Memory) -> str`
- [ ] Generate UUID if not provided
- [ ] Convert Memory to properties dict
- [ ] Create Cypher CREATE query
- [ ] Handle duplicate key errors
- [ ] Return memory ID
**Cypher Pattern**:
```cypher
CREATE (m:Memory {
id: $id,
type: $type,
title: $title,
content: $content,
tags: $tags,
importance: $importance,
created_at: $created_at,
updated_at: $updated_at
})
RETURN m.id
```
### 2.2 Implement get_memory()
**Tasks**:
- [ ] Implement `async def get_memory(self, memory_id: str, include_relationships: bool = True) -> Optional[Memory]`
- [ ] Query by memory ID
- [ ] Optionally fetch relationships
- [ ] Convert result to Memory object
- [ ] Return None if not found
### 2.3 Implement update_memory()
**Tasks**:
- [ ] Implement `async def update_memory(self, memory: Memory) -> bool`
- [ ] Use Cypher MATCH...SET pattern
- [ ] Update updated_at timestamp
- [ ] Return True on success, False if not found
### 2.4 Implement delete_memory()
**Tasks**:
- [ ] Implement `async def delete_memory(self, memory_id: str) -> bool`
- [ ] Delete memory node
- [ ] Delete all connected relationships
- [ ] Return True on success
### 2.5 Implement get_all_memories()
**Tasks**:
- [ ] Implement `async def get_all_memories(self, limit: int = 100, offset: int = 0) -> List[Memory]`
- [ ] Support pagination with SKIP and LIMIT
- [ ] Order by created_at descending
- [ ] Convert results to Memory objects
---
## Section 3: Relationship Methods
### 3.1 Implement create_relationship()
**Tasks**:
- [ ] Implement `async def create_relationship(self, relationship: Relationship) -> str`
- [ ] Validate source and target memories exist
- [ ] Create relationship with properties
- [ ] Handle duplicate relationship errors
- [ ] Return relationship ID
**Cypher Pattern**:
```cypher
MATCH (a:Memory {id: $from_id}), (b:Memory {id: $to_id})
CREATE (a)-[r:$type {
id: $id,
strength: $strength,
confidence: $confidence,
created_at: $created_at
}]->(b)
RETURN r.id
```
### 3.2 Implement get_relationships()
**Tasks**:
- [ ] Implement `async def get_relationships(self, memory_id: str, direction: str = "both", relationship_type: Optional[str] = None) -> List[Relationship]`
- [ ] Support direction: "outgoing", "incoming", "both"
- [ ] Optional filter by relationship type
- [ ] Convert results to Relationship objects
### 3.3 Implement delete_relationship()
**Tasks**:
- [ ] Implement `async def delete_relationship(self, from_id: str, to_id: str, relationship_type: str) -> bool`
- [ ] Delete specific relationship between nodes
- [ ] Return True on success
### 3.4 Implement get_related_memories()
**Tasks**:
- [ ] Implement `async def get_related_memories(self, memory_id: str, relationship_types: Optional[List[str]] = None, max_depth: int = 1) -> List[Memory]`
- [ ] Support multi-hop traversal
- [ ] Filter by relationship types
- [ ] Return unique memories
---
## Section 4: Search and Query Methods
### 4.1 Implement search_memories()
**Tasks**:
- [ ] Implement `async def search_memories(self, search_query: SearchQuery) -> List[Memory]`
- [ ] Support text search on title and content
- [ ] Support filtering by:
- [ ] Memory types
- [ ] Tags
- [ ] Date range
- [ ] Importance threshold
- [ ] Support pagination
- [ ] Support sorting
### 4.2 Implement recall_memories()
**Tasks**:
- [ ] Implement natural language recall if applicable
- [ ] Or delegate to search_memories with appropriate query
### 4.3 Complete Schema Initialization
**File**: `/Users/gregorydickson/claude-code-memory/src/memorygraph/backends/ladybugdb_backend.py:153-180`
**Tasks**:
- [ ] Add multi-tenant index support (check `Config.is_multi_tenant_mode()`)
- [ ] Add all indexes from FalkorDBLite:
- [ ] type index
- [ ] created_at index
- [ ] importance index
- [ ] confidence index (for relationships)
- [ ] Add proper error handling for constraint creation
- [ ] Add debug logging for schema operations
---
## Section 5: Testing
### 5.1 Fix Unit Test Mocking
**File**: `/Users/gregorydickson/claude-code-memory/tests/backends/test_ladybugdb_backend.py`
**Tasks**:
- [ ] Fix duplicate return value in `setup_mock_ladybug()` (line 68-74)
- [ ] Align mock result format with actual LadybugDB behavior
- [ ] Add tests for all new methods:
- [ ] store_memory tests
- [ ] get_memory tests
- [ ] update_memory tests
- [ ] delete_memory tests
- [ ] search_memories tests
- [ ] create_relationship tests
- [ ] get_relationships tests
- [ ] delete_relationship tests
- [ ] Add edge case tests:
- [ ] Empty results
- [ ] Not found errors
- [ ] Duplicate key errors
- [ ] Connection errors
### 5.2 Complete Integration Tests
**File**: `/Users/gregorydickson/claude-code-memory/tests/backends/test_ladybugdb_integration.py`
**Tasks**:
- [ ] Fix duplicate cleanup code (lines 60-62)
- [ ] Use unique table names to prevent test conflicts
- [ ] Add comprehensive integration tests:
- [ ] Full Memory CRUD lifecycle
- [ ] Relationship creation and traversal
- [ ] Search with various filters
- [ ] Multi-hop relationship queries
- [ ] Pagination
- [ ] Error conditions
### 5.3 Add Security Tests
**Tasks**:
- [ ] Test parameterized query handling
- [ ] Test injection attempt handling
- [ ] Test input validation
---
## Section 6: Documentation
### 6.1 Update Module Docstring
**File**: `/Users/gregorydickson/claude-code-memory/src/memorygraph/backends/ladybugdb_backend.py:1-6`
**Tasks**:
- [ ] Fix inaccurate comparison to Kuzu
- [ ] Document actual API compatibility
- [ ] Add security notes if parameterization not supported
- [ ] Document result format handling
### 6.2 Update README
**File**: `/Users/gregorydickson/claude-code-memory/README.md`
**Tasks**:
- [ ] Add LadybugDB to supported backends section
- [ ] Document installation: `pip install real-ladybug`
- [ ] Add configuration example
- [ ] Document environment variables
### 6.3 Add Backend-Specific Documentation
**File**: `/Users/gregorydickson/claude-code-memory/docs/backends/LADYBUGDB.md` (new)
**Tasks**:
- [ ] Create backend-specific documentation
- [ ] Document setup and configuration
- [ ] Document limitations vs other backends
- [ ] Add troubleshooting section
---
## Acceptance Criteria
### Functional
- [ ] All GraphBackend abstract methods implemented
- [ ] All Memory CRUD operations working
- [ ] All Relationship operations working
- [ ] Search with filters working
- [ ] Pagination working
- [ ] Multi-tenant mode supported (if enabled)
### Quality
- [ ] No SQL/Cypher injection vulnerabilities
- [ ] Consistent exception handling
- [ ] Type hints on all methods
- [ ] Docstrings on all public methods
- [ ] 90%+ test coverage for new code
### Testing
- [ ] All unit tests passing
- [ ] All integration tests passing (when real_ladybug installed)
- [ ] Tests properly skip when dependency missing
### Documentation
- [ ] README updated
- [ ] Module docstrings accurate
- [ ] Backend-specific docs created
---
## Success Metrics
| Metric | Target |
|--------|--------|
| Methods implemented | 100% of GraphBackend |
| Test coverage | 90%+ |
| Security issues | 0 |
| Integration tests | 15+ |
| Documentation | Complete |
---
## Dependencies
**External**:
- `real-ladybug` Python package
**Internal**:
- GraphBackend base class
- Memory and Relationship models
- Config for multi-tenant mode
---
## Reference Files
Use these as implementation references:
- `/Users/gregorydickson/claude-code-memory/src/memorygraph/backends/falkordblite_backend.py` - Primary reference
- `/Users/gregorydickson/claude-code-memory/src/memorygraph/backends/sqlite_backend.py` - SQL patterns
- `/Users/gregorydickson/claude-code-memory/src/memorygraph/backends/base.py` - Interface definition
---
## Notes for Coding Agent
1. **Start with Section 1** - Infrastructure must be fixed before methods can work
2. **Use FalkorDBLite as reference** - It's the most similar backend
3. **Test with real_ladybug if available** - Integration tests are critical
4. **Security first** - No string interpolation in queries
5. **Maintain consistency** - Follow patterns from other backends
---
**Last Updated**: 2025-12-08
**Status**: PENDING
**Next Step**: Complete WP16 and WP20 releases first
