Minotauris 0.6.8 is the biggest desktop experience update we have shipped so far. It improves the core UI, computer control, browser automation, bring-your-own-key support, memory, and long-running work.
Biggest desktop UI redesign so far
The desktop app now has a broader redesign across active work. The new inspector makes live runs easier to understand, Assembly is easier to inspect, and the canvas is tuned for longer sessions where users need to scan decisions, browser evidence, terminal output, memory, todos, and agent state.
The Assembly node has been rebuilt around the actual shape of group decision-making: members, discussion, ideas, votes, elapsed time, and verdict state are visible in a cleaner layout. The inspector now understands Assembly directly, showing phase, vote tally, proposals, recent discussion, and member reasoning.
Next generation computer use
This release improves desktop control so agents can act more reliably than raw coordinate clicking alone.
Desktop automation is still hard and some apps may still break. The important change is simple: agents now have a clearer path toward dependable local app control while still reporting failures honestly.
Better browser automation and visual evidence
Browser automation is stronger in isolated sessions. Agents can navigate, inspect pages, and capture screenshots when the selected model supports image input.
The steering and comment inputs now support local images for compatible agents. Pasted images remain local data URLs until the user sends them.
Bring-your-own-key, Azure, and smarter model choices
Minotauris 0.6.8 expands bring-your-own-key support with Azure OpenAI compatibility and improves model selection around real agent roles.
That makes work more efficient and more cost aware. Expensive reasoning can stay where it matters, while routine browser, coding, or research work can use the best model for the route you pick.
Smoother runs and stronger evidence checks
This update includes broad desktop optimization work aimed at smoother runs. The agent system now does more to preserve useful state, reduce waste, and keep long-running automation grounded in evidence.
New anti-hallucination methods improve longer automation runs by pushing agents toward inspectable evidence, explicit verification, persisted memory, and clearer reporting. The result is a system that is better at continuing work without drifting away from what actually happened.
Self-evolving system and efficient memory dreams
Minotauris is also getting better at remembering durable patterns from repeated work without turning memory into noise.
Together, these changes move Minotauris closer to a desktop app where agents can see, decide, remember, control tools, and keep users oriented through complex automation.