codexbar

A command-line tool for viewing usage limits, credits, and reset times for AI services such as Codex, Claude, Kimi, Gemini, Copilot, Cursor, and OpenRouter. It can read existing provider sessions and support multiple accounts.

In plain words
What is it for?
Use it to check provider usage in text or JSON, inspect a specific provider or account, and view available quotas using existing logins or configuration.
Why use it?
It gives a single way to check how much service capacity remains instead of opening each provider separately. It also shows when limits or credits are expected to reset.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/steipete/agent-scripts/codexbar
Any agent
npx skills add steipete/agent-scripts --skill codexbar
Clone the repo
git clone --depth 1 https://github.com/steipete/agent-scripts

Made for: Claude Code, Codex.

Per session 34 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 674 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00034 $0.00674
Opus 5 $0.00017 $0.00337
Sonnet 5 $0.00007 $0.00135
Haiku 4.5 $0.00003 $0.00067

Measured yesterday against content hash 1fcdb4cc0a96, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

codexbar scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/codexbar/SKILL.md · 63 lines

How it starts

The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CodexBar

Use this when asked about AI subscription usage, rate limits, remaining credits, or "how much Codex/Claude/Kimi do I have left".

CLI

  • Binary: codexbar (Homebrew)
  • Default invocation honors the in-app provider toggles:
codexbar usage
codexbar usage --format json
  • Specific provider (--provider accepts codex, claude, kimi, gemini, copilot, cursor, openrouter, and many more — see codexbar --help):
codexbar usage --provider codex --json
codexbar usage --provider claude --json
  • Multiple accounts: --account <label>, --account-index <n>, or --all-accounts (single provider only).

Auth model

Provider fetches ride existing sessions — OAuth caches, CLI credentials, or browser cookies. No secrets are passed on the command line.

  • Codex: OpenAI web dashboard via oauth; falls back to Codex CLI.
  • Claude: claude.ai API via browser session cookie; falls back to Claude CLI. "No Claude session key found in browser cookies" means log in to claude.ai in a supported browser (or rely on the CLI fallback).
  • Token-based providers read the CodexBar config file.

Dashboard endpoints

The menu bar app serves its dashboard on loopback. The page pulls /dashboard/v1/snapshot and /cost; there is no /api/* namespace, so those paths 404 rather than telling you that you guessed. The snapshot endpoint requires auth, so when the CLI is unavailable the quickest honest check is to open the dashboard in a browser.

usage --all-accounts polls every account serially and takes well over ninety seconds. If it produces zero bytes on both stdout and stderr, it is not slow, it is blocked: the CLI is a helper inside CodexBar.app, and a pending macOS Gatekeeper prompt stops it dead with no output. Screenshot the screen before debugging the command.

Interpreting output

  • Errors for providers that were never configured are noise, not failures — only report providers the user actually uses.
  • Report percentage left, pace vs. expected, and reset time; the pace line ("in deficit"/"in reserve") is the actionable bit.
  • Claude rows read claude-swap stored backups, which produces two confusing readings. The currently active slot shows as having no stored credentials, because the live login is an active profile and its backup only materialises when you switch away. And claude-swap keeps a last seen percentage that keeps rendering for accounts whose credentials are gone, so a dead account can read as healthy for days. A plausible percentage is not proof of a working credential; only refresh token yes in claude-swap list --token-status is.

Read the full file on GitHub · 63 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. yesterday First seen · 63 lines · 34 tokens per session scan A 1fcdb4cc0a96

Subscribe to this mod's changes

codexbar is a skill published in the GitHub repository steipete/agent-scripts (6,580 stars, last pushed 2d ago), licensed MIT. It adds 34 tokens to every session and 674 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens