Concepts
These are the building blocks behind every Maestrly workflow.
Provider and model
A provider operates the AI service; a model is one of the inference options that provider makes available. You connect your own provider account or API credentials. Requests go directly from Maestrly to the selected provider, whose terms, pricing, limits, and regional availability apply.
Project and chat
A project connects Maestrly to one or more folders or Git repositories on your machine. A chat is a working thread inside that project. It holds the request history, selected model, execution context, and the actions taken for a piece of work.
Agent and skill
An agent is a reusable working configuration: role, instructions, preferred model, and allowed capabilities. A skill is a focused procedure or body of guidance an agent can apply when a task calls for it. Use agents for stable behavior and skills for specialized, reusable know-how.
Tool and MCP
A tool lets an agent act or retrieve information. MCP (Model Context Protocol) is a standard way to expose tools and resources from local processes or remote services. A connected MCP server has its own permissions, credentials, network behavior, and availability.
Git execution mode
A chat can work in the project's current checkout or in an isolated Git worktree and branch. The current checkout is useful for direct collaboration; a worktree is safer for parallel or automated tasks. Isolation reduces collisions but does not replace review.
Local-first, not local-only
Maestrly keeps project coordination and local execution close to your machine. Network boundaries still matter:
- AI prompts, selected context, and responses travel directly between the app and your chosen provider for inference.
- Remote MCP servers and other connected services receive the data needed for calls you authorize.
- Maestrly has no account, hosted backend, telemetry, automatic diagnostic upload, or automatic update channel. Conversations, notes, settings, and usage records stay in the local profile.
Maestrly does not act as the inference provider or bundle model access.
Usage and responsibility
Maestrly is free, but providers and connected services may charge your own accounts. You control what is connected, what an agent may use, and whether a result is accepted, committed, merged, or deployed.
Next: Features overview.