feat: Redesign v2 resource endpoints to use entity IDs

Changed v2 resource endpoints to use entity IDs in URL paths instead of file
paths, with file_path moved to request bodies when needed. This is consistent
with v2's ID-first design principle.

New API Design:
- GET /v2/projects/{project_id}/resource/{entity_id} - Get content by entity ID
- POST /v2/projects/{project_id}/resource - Create resource (file_path + content in body)
- PUT /v2/projects/{project_id}/resource/{entity_id} - Update resource (content in body, optional file_path to move)

Key Changes:
- Uses integer entity IDs in URL paths (not file paths)
- File paths are in request/response bodies
- More RESTful: POST for create, PUT for update
- Supports moving files during update via optional file_path in request body
- Returns entity_id in all responses (ResourceResponse schema)

Request Schemas:
- CreateResourceRequest: {file_path: str, content: str}
- UpdateResourceRequest: {content: str, file_path?: str}
- ResourceResponse: {entity_id: int, file_path: str, checksum: str, ...}

Tests:
- Rewrote all 11 v2 resource tests for new ID-based API
- Tests cover create, read, update, move, path validation
- All tests passing with entity ID validation

Benefits:
- Stable entity references (IDs don't change when files move)
- Cleaner URL structure without path encoding issues
- Consistent with other v2 endpoints (knowledge, memory, etc.)
- Better separation of concerns (ID for routing, path for storage)

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Joe P <joe@basicmemory.com>

Author: jope-bm