# About

Maestria is portable AI engineering praxis: a set of free, open-source plugins that install a complete working methodology into AI coding agents, instead of asking you to memorize one. Once installed through a single `npx maestria install <platform>` command, an agent gains a dispatch-only orchestrator supported by seven specialists - adventurer (reconnaissance), architect (approach and trade-offs), builder (atomic implementation), diagnose (root-cause analysis), planner (phased plans), reviewer (independent review), and writer (documentation). The same methodology then runs natively on OpenCode, Kimi Code, Pi, Hermes, Claude Code, Codex CLI, Cursor, prime-agent, and OMP.

Two patterns anchor the methodology. **Pipeline composition** routes work through focused stages - thinker, worker, verifier - instead of asking one context to do everything. The **maker/checker split** keeps production and approval separate: the builder produces, and the reviewer sees the work fresh and independent, so no agent approves its own output. You can read how these fit together in [How It Works](/core/how-it-works/), [Pipeline & Roles](/core/pipeline/), and the [Specialist Reference](/core/agents/), or start with [When to Use Maestria](/core/when-to-use/) to decide between direct execution, a specialist, or the full pipeline.

Maestria is open source under the [MIT license](https://github.com/agustinusnathaniel/maestria/blob/main/LICENSE) and is developed and maintained by Agustinus Nathaniel ([agustinusnathaniel](https://github.com/agustinusnathaniel)), with contributions welcome through the [GitHub repository](https://github.com/agustinusnathaniel/maestria). To reach the maintainer, use the channels on the [Contact](/contact/) page; for details on how this site handles data, see [Privacy](/privacy/).