​

Changelog

All notable changes to OmniCoreAgent are documented here.

---

​

[0.1.18] - 2025-06-19

​

Changed

• BREAKING CHANGE: Removed local ChromaDB support for performance optimization
  • Local ChromaDB fallback system has been completely removed
  • Users must now explicitly configure a vector database provider
  • No more automatic fallback to local storage
  • This eliminates the 81+ second startup delay from ChromaDB Rust bindings
• Performance Improvement: Vector database modules now load only when explicitly needed
  • Lazy loading prevents unnecessary imports during package initialization
  • Faster startup when vector DB is disabled
  • Better resource management

​

Added

• Required Configuration: Users must set OMNI_MEMORY_PROVIDER to one of:
  • qdrant-remote (recommended default)
  • chroma-remote
  • chroma-cloud
• Clear Error Messages: Better feedback when vector DB configuration is missing

​

Removed

• Local ChromaDB persistence and warmup system
• Automatic fallback mechanisms
• Module-level vector DB initialization
• ChromaDB local client type

​

Migration Guide

• Before: ENABLE_VECTOR_DB=true would automatically use local ChromaDB
• After: Must set both ENABLE_VECTOR_DB=true AND OMNI_MEMORY_PROVIDER=qdrant-remote (or other provider)
• Impact: Faster startup, but requires explicit configuration

---

​

[0.1.17] - 2025-05-28

​

Added

• OAuth Authentication Support:
  • OAuth 2.0 authentication flow for MCP servers
  • Support for multiple authentication methods (OAuth 2.0, Bearer token, Custom headers)
  • Flexible authentication configuration in server settings
  • Secure credential management
• Enhanced Server Configuration:
  • Updated server configuration format to support OAuth
  • Improved server connection security

​

Changed

• Updated server configuration examples to include OAuth support
• Enhanced documentation for authentication methods

​

Fixed

• Improved authentication error handling
• Updated configuration validation for authentication methods

---

​

[0.1.16] - 2025-05-16

​

Added

• Streamable HTTP Transport: New transport protocol with efficient data streaming, configurable timeouts, and header support
• Dynamic Server Management:
  • /add_servers:<config.json> — Add one or more servers
  • /remove_server:<server_name> — Remove servers
  • Real-time server capability updates after changes
• Enhanced server configuration with streamable HTTP examples

​

Changed

• Updated README with new server management commands
• Improved documentation for transport protocols

​

Fixed

• Improved server connection handling
• Enhanced error messages for server management commands

---

​

[0.1.15] - 2025-05-05

​

Added

• Token & Usage Management:
  • /api_stats command for token/request tracking
  • Configurable limits for total requests and token usage
  • Configurable tool call timeout and max steps
• Developer Integration Enhancements:
  • FastAPI example for building custom API servers
  • Minimal code snippets for custom MCP client integration
• FastAPI API Documentation: /chat/agent_chat endpoint docs
• Environment Variables Reference: Full table in README

​

Changed

• Updated server configuration examples
• Improved README structure

​

Fixed

• Corrected typos in documentation
• Improved code example consistency

---

​

[0.1.14] - 2025-04-18

​

Added

• DeepSeek model integration with full tool execution support
• Orchestrator Agent Mode:
  • Advanced planning for complex multi-step tasks
  • Strategic delegation across multiple MCP servers
  • Parallel task execution capabilities
  • Dynamic resource allocation
  • Real-time progress monitoring
• Client-Side Sampling Support: Dynamic sampling configuration
• Chat History File Storage: Save/load conversations, persistent history

​

Changed

• Enhanced Mode System with three distinct modes (Chat, Autonomous, Orchestrator)
• Updated AI model integration documentation

​

Fixed

• Improved mode switching reliability
• Enhanced chat history persistence

---

​

[0.1.13] - 2025-04-14

​

Added

• Gemini model integration with full tool execution support
• Redis-powered memory persistence:
  • Conversation history tracking
  • State management across sessions
  • Multi-server memory synchronization
• Agentic Mode capabilities:
  • Autonomous task execution
  • Advanced reasoning and decision-making
  • Complex task decomposition
• Advanced prompt features: Dynamic discovery, JSON/key-value formats, nested arguments
• Comprehensive troubleshooting guide
• Detailed architecture documentation
• Advanced server configuration examples

​

Changed

• Enhanced installation process with UV package manager
• Improved development quick start guide
• Updated server configuration format for multiple LLM providers

​

Fixed

• Standardized command formatting
• Improved code block consistency

---

​

[0.1.1] - 2025-03-27

​

Added

• Comprehensive Security & Privacy section
• Detailed Model Support section (OpenAI, OpenRouter, Groq)
• Structured Testing section with coverage reporting
• Support for additional LLM providers (OpenRouter, Groq)

​

Changed

• Improved AI-Powered Intelligence section with multi-provider support
• Enhanced Server Configuration Examples
• Updated Python version requirement from 3.12+ to 3.10+
• Changed from OPENAI_API_KEY to LLM_API_KEY for broader provider support

​

Fixed

• Various typos and formatting issues
• Fixed typo in “client” command (was “cient”)

​

Removed

• Redundant security information
• Specific OpenAI model references in favor of provider-agnostic examples

---

​

[0.1.0] - 2025-03-21

• Initial release
• Basic MCP server integration
• OpenAI model support
• Core CLI functionality