v0.3.1

Official Langflow Version of Agentics

What's Changed

• Semantic operators fixes by @D3f0 in #136
• Semantic operators by @D3f0 in #137

Full Changelog: v0.3.0...v0.3.1

0.3.0 - Semantic Operators & Documentation

🎯 Key Features

Semantic Operators

• sem_map: Apply semantic transformations across datasets
• sem_filter: Semantically filter data based on LLM-powered predicates
• sem_agg: Aggregate semantic information across collections

Documentation & Architecture

• Comprehensive documentation improvements for semantic operators
• Refactored LLM connections module for better architecture
• Enhanced server_url parameter support for flexible LLM configuration
• Updated README with semantic operators integration guide

Bug Fixes & Improvements

• Fixed undefined variable in PydanticTransducerVLLM executor
• Corrected dependency references in pyproject.toml
• Fixed package imports in async_executor.py
• Improved code formatting and pre-commit compliance
• Enhanced module organization and separation of concerns

📦 Installation

Install the latest version:

pip install agentics-py==0.3.0

📖 Documentation

• Getting Started: https://github.com/IBM/Agentics/blob/main/docs/getting_started.md
• API Reference: https://github.com/IBM/Agentics/blob/main/docs

---

Release Date: February 12, 2026
Status: Latest Stable Release

v0.2.4

What's Changed

• Relax pandas version constraint to >=2 by @D3f0 in #126

Full Changelog: v0.2.3...v0.2.4

v0.2.3

What's Changed

• Add semantic operators module by @D3f0 in #124

Full Changelog: v0.2.2...v0.2.3

v0.2.2

Changes

Core Dependencies

• Added duckdb>=1.4.3 for efficient data querying
• Added pandas>=2.3.3 for enhanced data manipulation
• Added async-lru>=2.0.5 for async caching capabilities

Type System Enhancements ( src/agentics/core/atype.py )

• Added comprehensive docstring to pydantic_model_from_dict() function explaining:
• Dynamic Pydantic model creation from sample dictionaries
• Type inference and field name normalization
• Usage examples and edge cases
• Renamed composition fields from left/right to target/source for semantic clarity:
• target : The destination type in a composition
• source : The source type in a composition
• Removed debug print statement from create_pydantic_model()

Default Types ( src/agentics/core/default_types.py )

• Added Abool type: A new semantic type for boolean values with type validation
• Ensures only boolean values are accepted during construction
• Follows the same pattern as existing Astr type

Transduction Improvements ( src/agentics/core/transducible_functions.py )

• Enhanced Transduce class with string representation methods:
• Added str() : Human-readable string output
• Added repr() : Developer-friendly representation
• Added _one_to_str() : Handles Pydantic models, dicts, lists, and generic types
• Added new parameters to transducible() decorator:
• transduce_fields : List of specific fields to transduce
• prompt_template : Custom prompt template for transduction
• Fixed transduction return type: Now returns TransductionResult consistently
• Fixed generate_prototypical_instances() : Updated field definition to use optional list

Core AG Enhancements ( src/agentics/core/agentics.py )

• Improved batch processing:
• Transduction now chunks inputs using amap_batch_size for better memory management
• Processes chunks sequentially to handle large datasets
• Fixed state merging logic:
• Now filters output fields against allowed model fields (Pydantic v2 compatibility)
• Prevents errors from extra fields in merged states
• Validates AG lengths before merging
• Enhanced explanation generation:
• Improved prompt instructions for better semantic understanding
• Renamed explanation variables for clarity ( left/right → target/source )
• Added confidence scoring guidance to LLM
• Improved compose_states() method:
• Changed from nested loops to paired zip() iteration (matching AG lengths)
• Fixed semantic naming: left/right → target/source
• More efficient and semantically correct composition
• Better error handling:
• Added validation in merge_states() to check length compatibility
• Meaningful error messages for debugging
• Truncated display names: Limited AG names to 30 characters in progress descriptions

v0.2.2a2 Preview - Semantic operators improvements

Changes

This alpha release includes significant improvements to the Agentics framework's type composition system and transduction capabilities.

Key Features

• Enhanced Type Composition: Renamed composition fields from left/right to target/source for better semantic clarity
• New Abool Type: Added semantic type for boolean values with validation
• Improved Batch Processing: Transduction now chunks inputs for better memory management
• Better State Merging: Fixed Pydantic v2 compatibility with proper field filtering
• Enhanced Explanations: Improved LLM instructions for more meaningful semantic transductions
• New Dependencies: Added duckdb, pandas, and async-lru for improved data handling

Technical Improvements

• Chunked transduction prevents memory overload with large datasets
• Better error handling with meaningful validation messages
• Improved string representations for Transduce objects
• Pre-commit fixes applied

For detailed information, see PR #115

v0.2.2a1 Pre-release

Pre-release version v0.2.2a1

LiteLLM proxy support and LLM connection debugging

What's Changed

• Agentics2.0 - Updated Documentation from AG 1.0 version by @gliozzo in #102
• Fix: correct undefined variable 'state' to 'input' in PydanticTransducerVLLM.execute() and dependencies imports by @mnlscn in #104
• LiteLLM AI Gateway support by @D3f0 in #101

New Contributors

• @mnlscn made their first contribution in #104

Full Changelog: v0.2.0...v0.2.1

v0.2.1a4 Preview - LiteLLM Proxy & Provider Visibility (Updated)

Preview Release: LiteLLM Proxy and LLM Provider Visibility

✨ Features

• LiteLLM Proxy Support: Full configuration with LITELLM_PROXY_URL, LITELLM_PROXY_API_KEY, and LITELLM_PROXY_MODEL
  • Model name validation (must start with litellm_proxy/)
  • Logger warnings for invalid configurations
• LLM Provider Visibility: New show-llms CLI command with rich tables
  • Display all available LLMs with their providers and models
  • Show environment variables used for each LLM
  • Group LLM aliases with instance vs alias count
  • Highlight active LLM in footer
• Environment Variable Tracking: Each LLM tracks which env vars were used for configuration
• Public API: get_available_llms() is now public for external use

🔧 Improvements

• Provider-specific parameter validation (Claude doesn't support both temperature + top_p)
• Env file update helper with change detection
• New invoke task for selecting LiteLLM Proxy models with fzf
• Updated .env_sample with comprehensive LiteLLM documentation
• Hunter added for development tracing

🐛 Bug Fixes

• Fix Claude model compatibility (remove top_p parameter conflict)
• Handle dynamic versioning from uvx

📦 Installation

pip install agentics-py==0.2.1a4

🚀 Usage

# List all available LLMs
show-llms

# Interactively select a LiteLLM Proxy model
uv run invoke change-litlellm-env-model

# Or with a query filter
uv run invoke change-litlellm-env-model --query gpt

📋 Squashed Commits

• feat: Add LiteLLM Proxy support and LLM provider visibility tools
• fix: Add provider-specific parameter validation for Claude models
• chore: Add hunter for tracing code in dev group

v0.2.1a3 Preview - LiteLLM Proxy Support

Preview Release: LiteLLM Proxy and LLM Provider Visibility

Features

• LiteLLM Proxy Support: Full configuration with LITELLM_PROXY_* environment variables and model validation
• LLM Provider Visibility: New show-llms CLI command displaying all configured LLMs in rich tables
• Environment Variable Tracking: Each LLM now tracks which env vars were used for configuration
• Alias Grouping: LLM aliases are grouped together with instance vs alias count display
• Provider-Specific Validation: Claude models automatically handled (no temperature + top_p conflict)
• Public API: get_available_llms() is now public for external use

Improvements

• Add env file update helper with change detection
• Make LLM configuration more transparent and debuggable
• Support 100+ LLM providers through LiteLLM integration

Bug Fixes

• Fix Claude model compatibility (temperature + top_p conflict)

Installation

pip install agentics-py==0.2.1a3

Usage

# List available LLMs
show-llms

# Select a model for LiteLLM Proxy
uv run invoke change-litlellm-env-model