- Added live timeout countdown badges, redesigned shell tool cards, and implemented hard budgets for subagents to prevent hangs.
- Fixed issues with tools getting stuck, improved error handling for invalid paths, and addressed UI freezes from high-frequency events.
- Changed default timeouts for Todo operations and tightened read timeouts to enhance performance consistency.
- Added per-tool hard timeouts and abort-signal support for long-running tools to enhance responsiveness.
- Introduced configurable timeout settings in the agent's configuration.
- Implemented GPU-accelerated local embeddings with fallback options, along with detailed indexing page information.
- Expanded ignored directories for tools and indexing to improve performance and efficiency.
- Added indexing toggle in settings, persistent semantic index, incremental re-indexing, real-time auto-indexing, and context size dropdown for models.
- Changed smart conversation summarization triggers, subagent history management, multitask handling, and stop/cancel behavior.
- Fixed issues with stuck subagent spinners, orphaned shell processes, and context ring defaults.
- Updated version in package.json to 0.0.3.
- Added per-workspace conversations with automatic migration from global conversations.
- Implemented auto-loading of GGUF models with a loading card in chat.
- Enhanced llama.cpp server to use random free ports with retry on bind failure.
- Improved composer dropdown positioning and shared model selection across composers.
- Fixed production error related to missing '@huggingface/hub' package.
- Removed MCP tool marketplace.
- Updated version in package.json to 0.0.2.
Pawan Osman