User Attribution for Automation Runs, Semantic Explorer Sorting, and Migration Improvements

Comprehensive user tracking for workflow and agent executions:

Audit Trail - Record which user triggered each automation run for better accountability and analytics
User Context Threading - User information flows through the entire execution context across all workflow and agent execution paths
Visual Attribution - UI displays user avatars and names for each run in the run selection dropdown
Database Schema - Added nullable user_id field to runs table with proper foreign key constraints and indexing

This feature enables better governance, user analytics, and execution attribution across the platform.

Manual sort criteria management for semantic queries:

Sort Order Control - Add, update, and remove sort criteria directly in the semantic explorer interface
New UI Components - SortsSection and SortRow components for intuitive sort management
Query Integration - SemanticQueryOrder type support throughout the query building pipeline
State Management - Extended hooks to handle sort operations seamlessly

Users can now manually define sorting behavior for their semantic queries, providing more control over result ordering.

Semantic Query Error Tracking - Added three distinct error types (validation_error, sql_generation_error, and error) for granular error reporting throughout the semantic query lifecycle
Incremental Error States - Refactored artifact flow to emit error states progressively, enabling better user feedback in the UI
Improved Error Visibility - Error fields added to semantic query artifacts for comprehensive error tracking from validation through execution

Foreign Key Constraint Resolution - Fixed foreign key constraint issues in SQLite migrations by recreating tables with proper schema
Optional User ID - Refactored user_id field from non-nullable UUID (using Uuid::nil() placeholder) to Option<uuid::Uuid> (using None for absence) for cleaner semantics
Migration Compatibility - Fixed INSERT statement compatibility by using explicit column lists instead of SELECT *, NULL pattern
Semantic Query Validation - Introduced SemanticQueryValidation enum for better handling of valid and invalid query states
JSON Schema Simplification - Custom JSON schema implementation for SemanticFilter compatible with OpenAI specifications

Latest Versions - Updated Rust and all dependencies to latest versions for improved performance and security