borawong/AiMaMi

What it solves

AiMaMi eliminates the need to manually edit complex configuration files (TOML, JSON, and SQLite) to manage OpenAI Codex. It solves the friction of switching accounts, handling quota exhaustion, configuring third-party model relays, and cleaning up local session threads.

How it works

Built with Tauri 2, React, and Rust, the application acts as a native desktop companion that reads and writes Codex data stored locally in the ~/.codex directory. It provides a graphical user interface for high-frequency workflows and includes a local proxy to forward relay models into Codex Desktop.

Who it’s for

Users of OpenAI Codex who want a streamlined way to manage their accounts, routing settings, and system maintenance without risking manual file corruption.

Highlights

• Account & Quota Management: Easy account switching and automatic fallback when quotas are exhausted.
• Smart Routing: Integration of relay models while maintaining resumable historical threads.
• Session & Data Cleanup: Tools to inspect, analyze, and bulk-clean local threads and rebuild registries.
• MCP & Skills Management: UI-based lifecycle management for MCP entries and Skills with backup/restore capabilities.
• System Maintenance: Built-in diagnostics, force-quit options for Codex, and configuration fix tools.