[Release] crewaiInc/crewAI 1.14.7a1: 1.14.7a1

crewAI 1.14.7a1 introduces native Snowflake Cortex LLM provider support and crew trained agents file functionality. The release includes integration guides for Databricks and Snowflake, fixes critical CLI and file input issues, and improves import performance through lazy-loading. Key refactoring splits flow.py into modular components while addressing tool result handling and multi-source listener reliability.

Key Points

• •Native Snowflake Cortex LLM provider integration enables direct LLM calls within Snowflake environment
• •Crew trained agents file support allows persistence and reuse of trained agent configurations
• •CLI fix restores [project.scripts] in crewai package to resolve UV tool installation failures
• •File input reliability improvements resolve inconsistencies in file handling across workflows
• •Snowflake Claude tool result histories now complete with proper stringified tool call handling
• •Multi-source or_ listeners re-armed across router-driven cycles for improved event handling
• •Import speed optimization via lazy-loading of docling imports reduces initial load time
• •Flow.py refactored into DSL, definition, and runtime modules for better code organization
• •New integration guides for Databricks and Snowflake enable faster adoption of data platform features

Workflow Diagram

Concepts