Maestro orchestration

A single chat that plans, retrieves, edits files, runs tools and schedules follow-ups. BYOK any model — local or cloud — and switch mid-conversation.

What it is

Maestro is the central AI surface in osFoundry. Every conversation is an orchestrator that can decompose your request, pick the right model for each step, run tools against your data, edit files with diffs you approve, and schedule itself to follow up. Local models and cloud APIs live side by side — switch mid-conversation by typing a command.

You stay in control: every tool call is sandboxed, every file edit is reviewable, every long-running plan asks for approval first. When the model gets it wrong, you roll back. When it gets it right, you save the prompt to your library and share it with the team.

Key capabilities

Plan → retrieve → tool → edit, all in one chat
BYOK any model — switch local / cloud mid-conversation
Tool approval, plan mode and rollback built in
Versioned prompts and pipelines, shareable across workspaces
Memory groups keep context across sessions and projects

Use cases

Engineers using one chat to research, code, deploy and document
Ops teams running scheduled multi-step workflows
Founders mixing local privacy with cloud capability for cost control