Codex Manager
Local-first desktop control center for OpenAI Codex — sessions, models, tokens, accounts, providers, and AGENTS.md.
Install
git clone https://github.com/ewsun22/codex-managerInstall command taken from the project's README or registry entry. Always review code before running it.
README excerpt
Read full README on GitHub →Codex Manager
English | 简体中文
Your local control center for OpenAI Codex.
Codex Manager is an unofficial, local-first desktop app for understanding and managing Codex usage on your own machine: sessions, models, reasoning effort, tokens, accounts, providers, projects, and AGENTS.md files in one place.
Download for macOS (Apple silicon) →
Download
Download the latest signed macOS release · Release assets and checksums · All releases
The latest non-draft, non-prerelease GitHub Release is the stable macOS arm64 download source of truth. Each version must independently pass Developer ID signing, Apple notarization, stapling, and public asset verification; verify its published assets and checksums when your workflow requires it.
If Codex Manager makes your workflow easier, consider starring the repository to follow future releases.
Known issue in v0.6.6: upgrading a database near 512 MiB can fail at migration 14 and prevent startup. The v0.6.7 source fixes this with a fixed 1 GiB database budget. Existing databases must not be deleted as a workaround. See diagnosis and validation.
Why Codex Manager?
As Codex usage grows, useful context is spread across the CLI, local rollout files, configuration, account state, and project instructions. It becomes difficult to answer which model and reasoning effort a task used, how many tokens were observed, which projects have an AGENTS.md, and which account or provider is active.
Codex Manager brings those views into a single local desktop control center. Observed data is labeled by source, missing values remain unavailable, and estimates are not presented as provider bills.
Features
Observe
- Browse sessions, turns, model interactions, models, providers, reasoning effort, timing, and available token categories.
- Inspect input, output, cached read/write, and reasoning output tokens with source/provenance.
- Filter and cursor-page activity; open detail views without persisting message or reasoning text.
- Estimate API-equivalent cost from a versioned pricing catalog. This is not ChatGPT subscription billing or a provider invoice.
- The overview uses a compact layout with one recent-activity row; the homepage API-equivalent estimate is rounded to two decimal places while coverage semantics remain unchanged.
Manage
Similar plugs
CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies
Bash is all you need - A nano claude code–like 「agent harness」, built from 0 to 1
Documented system prompts from Anthropic - Claude Fable 5.1, Opus 5.5, Claude Design, Claude Code. OpenAI - ChatGPT GPT-6-Astra, Codex. Google - Gemini 3.8 Flash, 3.1 Pro,…
Wrap Antigravity, ChatGPT Codex, Claude Code, Grok Build, Muse Code, Devin as an OpenAI/Gemini/Claude/Codex compatible API service, allowing you to enjoy the free Gemini…