Skip to content
PrumoPrumo
    Searches every heading in the documentation.

    The conventions live in the project

    A generated project carries a .prumo/ folder. It is the reason Prumo is not a scaffolder: the code arrives with the rules it follows, written down, in the repository where the work happens.

    Nine areas, and only the ones that apply

    Only core ships unconditionally, because everything in it is true of every type. The rest follow the answers, since advice about a platform the project does not have is worse than no advice at all.

    • core/Always
    • api/Only when the answers call for it
    • database/Only when the answers call for it
    • client/Only when the answers call for it
    • web/Only when the answers call for it
    • mobile/Only when the answers call for it
    • site/Only when the answers call for it
    • monorepo/Only when the answers call for it
    • multi-tenancy/Only when the answers call for it
    Read the documents

    Not tied to one assistant

    .prumo/ is the single source of truth and stays neutral. The scaffolder writes the entry file each tool loads, and nothing more: AGENTS.md is the real pointer, read by Codex, Cursor, Copilot, Gemini CLI and others, and CLAUDE.md is an import of it. No generated skills, subagents or slash commands, because those change every quarter.

    Then they are yours

    Once written, those documents belong to the team. Prumo does not come back and rewrite them. Updating a project’s .prumo/ in place is deliberately left until after 1.0.