feat: updarte

Fix formatter output to use consistent green color format

fix: add UTF-8 support and project format compatibility

chore: remove temporary test files

fix: use proper box drawing characters for tool output

feat: implement enhanced tool output formatting

docs: update idle screen example to reflect current state

feat: updarte

feat: updarte

feat: updarte

chore: update .gitignore to exclude screen captures

refactor: remove obsolete test files

feat: add test utilities and helper functions

feat: implement BDD-style E2E testing framework

feat: add comprehensive testing infrastructure for UI state validation

Merge: accept local changes and cleanup tests

refactor: remove obsolete test files and add e2e framework

refactor: update logging calls to use standard levels

feat: refactor logging system with standard levels

Fix context_renderer nil error on UI reopen

fix: Reinitialize event handler dependencies when showing interface

fix: Force reset session state when closing interface

fix: Properly display current session when reopening interface

fix: Prevent sessions from hanging in 'Waiting' state

enhance: Kill session also stops current request

feat: Add 'K' keybinding to help panel

feat: Add 'K' keybinding to kill current session

improve: Always refresh context when showing multi-window

fix: Context output not appearing after reopening multi-window

refactor: Remove API Configuration from Sessions mode

refactor: Clean up unused test files and enhance multi-window functionality

docs: Update optimization context with current state analysis

feat: Add real-time timer updates with color highlighting

feat: Add session and task timers with automatic timeout handling

Migrate from API.generate_code to requestLLM interface\n\n- Updated all test files to use requestLLM instead of API.generate_code\n- Updated example files to demonstrate the new requestLLM interface\n- Updated documentation to reflect the new preferred interface\n- Modified init.lua to deprecate direct export of generate_code\n- Created migration test to verify both interfaces work correctly\n- Maintained backward compatibility with API.generate_code in API v2\n- Added note about deprecation of generate_code in favor of requestLLM

feat: Add provider information to stats UI display

refactor: Migrate from api.lua to api_v2 with LLM abstraction layer

feat: Implement LLM abstraction layer with provider support

refactor: reduce coupling and increase cohesion in multi_window module

feat: add test

refactor: optimize project

feat: implement strategy pattern for single-use tools execution

test: add mock

feat: colored state

feat: add nui layout

Добавлены тесты для всех инструментов в отдельную директорию tests/tools/

Исправлен режим V в multi_window: теперь исключает DEBUG контент и правильно отображает llmContent

Enhancement: Add detailed context output for tool executions and LLM responses

Feature: Add stop_processing tool to allow LLM to terminate sessions

Refactor: Optimize test suite by consolidating redundant tests

Cleanup: Remove vim.notify from API setup to reduce UI clutter

Fix: Ensure stats show correct model information immediately

Fix: Ensure stats information is displayed immediately in UI

Fix: Show stats information immediately when UI is displayed

feat: add nui layout

Fix queue module reference in session manager

Reorganize project structure: Move utility files to lib/utils and update module references

Refactor configuration structure and remove outdated refactor documentation

docs: add module reorganization documentation

docs: add final refactor summary

refactor: remove unused modules

docs: mark refactor agent task as complete

docs: update refactor documentation with latest changes

refactor: remove unused and outdated files

feat: add nui layout

test: add bdd

Fix multi-window interface issues

Align multi-window implementation with window.ctx specification

test: add bdd

refactor: multi window in separate modules