minor tweak to switch LLM provider from Ollama to Gemini

Configured Ollaman and improved llm config functionality: now the default is ollama, the fallback is gemini 2.5 flash lite, and these can be changed modifying just on line in the yaml config file

Remove PNG files from git tracking

fixed caching

Merge branch 'omirl_tool'

refactor: Complete OMIRL architecture cleanup and modernization

Fix LLM configuration and OMIRL comune filtering

Merge omirl_refactor: Complete OMIRL refactoring with livelli_idrometrici

feat: Add livelli_idrometrici task with unified adapter-level caching

Fix task output and test infrastructure

feat: Complete OMIRL configuration refactoring - eliminate hardcoded values

feat: remove mode parameter from OMIRL tool - implement direct task-based routing

feat: Complete agent architecture refactoring and testing framework

🎯 Complete OMIRL web services refactoring with workflow enhancement

Phase 4: Implement LLM-based summarization node

🧹 Major cleanup: Remove obsolete code, modernize task-agnostic architecture

docs: Update TODO for next development session

feat: Complete OMIRL Phase 1 refactor - YAML-based validation architecture

refactor: Update OMIRL specification with new lean input/output schema

fix: resolve service validation and timeout issues in agent workflow

LLM router and summarization currently working. Fixed agent initialization issues and basic llm-based summarization service

Fix missing get_llm_config import and add comprehensive debugging

Add comprehensive environment debugging for HF Spaces

Merge branch 'main' of https://github.com/cionny/promptaid-operations

Fix: Implement lazy agent initialization and improve environment loading

Merge pull request #1 from cionny/simple-llm-router

Comprehensive update: LLM Router implementation and Docker fixes

Fix Docker Playwright installation and Streamlit PATH issues

Updated the adapter and the agent workflow (node, state, registry etc.). Updated web app, currently supporting omirl scraping

hard coded the 12 sensor types that were wrong in the table scraper and fixed test that were getting stuck because there was no browser session clean up

Issues with hf deployment, fixed playright dependencies

Also updated requirements and docker file. App works locally, need to deploy on hf

Set up of the project structure: langgraph and tool based agent. Implemented first tool and task: omirl_snapshot_summary adapter can scrape data from sensorstable omirl. Next step: add tool to agentic framework and deploy to hugging face. Next step: scrape data based on sensor filter.

Add HF space metadata config in readme

Initial commit with docker image ready for Hugging Face Spaces deployment