agentic_architecture
A production-ready Flutter package for building intelligent agent-based applications with support for single agents, Mixture of Experts (MOE), and pluggable LLM providers.
What this package is like to depend on
Last release 10 months ago
15 Oct 2025
Too new to tell
only 2 dated releases
Nearly every release is documented
notes for 2 of 2 stable releases
Nothing withdrawn
no release was ever pulled
10 months old
2 releases · first in 2025
2 releases in the last 12 months
see the full history below
Release timeline
2 releases · Oct 2025 to Oct 2025
2026
Releases
latest 2-
0.1.115 Oct 2025Release notes
Open source →Fixed:
- Replaced print() with proper logging framework
- Resolved pub.dev static analysis warnings
- Added AgenticLogger for production-safe logging
No breaking changes - patch release
Release notes
Open source →Fixed
- Replaced all
print()statements in library code with proper logging framework - Created
AgenticLoggerusingdart:developerfor production-safe logging - Resolved pub.dev static analysis warnings about
avoid_print - All logging now controlled by
enableLoggingflag inAgentConfig
Technical Details
- Added
lib/src/core/logger.dartwith production-ready logging - Updated 6 files to use new logger instead of print
- No breaking changes - all functionality preserved
- All 32 tests still passing
-
0.1.011 Oct 2025Release notes
Open source →🎉 First stable release of agentic_architecture
Key Features:
- Complete agent system with tools
- MOE (Mixture of Experts) support
- 4 LLM providers (DeepSeek, OpenAI, Claude, Gemini)
- Production-ready with full test coverage
- Comprehensive documentation and examples
Package Stats:
- 4,615 lines of code
- 32 passing tests
- 53 KB package size
- 0 pub.dev warnings
- MIT Licensed
Release notes
Open source →Added
- Initial release of Agentic Architecture package
- Core agent abstraction with
Agentbase class - Tool system with
Toolinterface - LLM provider abstraction with
LLMProviderbase class - Agent coordinator for orchestrating agents
- Multiple LLM provider implementations:
- DeepSeek
- OpenAI
- Claude (Anthropic)
- Gemini (Google)
- Mixture of Experts (MOE) system:
- Lead agent orchestration
- Expert agent specialization
- Intelligent expert routing
- Response synthesis
- Delegation strategies:
- Single expert
- Parallel execution
- Sequential execution
- Intelligent (LLM-based)
- Conversation memory support
- Response formatting service
- Educational features:
- Token tracking
- Cost analysis
- Logging service
- Example tools:
- Echo tool
- Calculator tool
- Weather tool
- Example applications
- Comprehensive documentation