SUMMARY AI要約 by gpt-5-mini
aitok is a lightweight offline CLI tool to summarize local token usage and estimated USD cost for Claude Code, Codex, and Gemini CLI runs. It parses local tool logs (never uploads data or reads API keys) to produce counts, token totals, cache stats, cost estimates, and per-model/provider breakdowns. Who uses it: developers and operators who run local AI CLIs or want an offline audit of usage/cost across sessions and tools. Key features: - Offline summaries and reports (summary, report, pricing audit, budget check, tui, doctor) - Thread/session inclusion with titles, token/event details, and resume support for local sessions - Filters/grouping by tool, model, provider, day, cwd; JSON output for automation - Built-in pricing catalog plus configurable ~/.aitok/pricing.json and interactive configure flow - Provider attribution rules for Codex based on local evidence - Install via Homebrew, shell script, or go; optional version checks and upgrade helpers - Budget check returns nonzero exit on limit exceed while emitting full JSON payload
使用AIコーディングツール
このリポジトリには以下のAIコーディングツールの設定ファイルが置かれており、開発時にAIアシスタントが利用された痕跡があります。
DETECTED 検出されたAIスタック
このRepoのdescription / GitHub topics / READMEのAI要約に出現したAI関連キーワードをカテゴリ別に表示。各バッジは該当カテゴリの詳細ランキングへリンクします。
使用言語(バイト数比)
オーナー情報
日付
| GitHub作成日 | 2026-05-08 |
| 最終Push | 2026-05-09 |
| 当サイト初検出 | 2026-05-09 |
| 最終取得 | 2026-05-09 14:18 |
類似Repo (同じ言語のAI関連Repo)
Claude Code Go rewrite project
klampatech/ralph-cliAll-in-one Ralph Loop + CLI — single static Go binary that bundles ClaytonFarr/ralph-playbook and orchestrates Claude Code for autonomous spec→code loops.
felipematheus1337/InsightFlow-AIPlataforma corporativa para coleta, análise e consolidação de informações operacionais. Funcionários registram atividades, solicitações e dados de negócio, enquanto uma camada de IA gera relatórios executivos automatizados enviados aos gestores, permitindo decisões rápidas e baseadas em evidências.
martinhg/capiko-ai boldfield/agentaskAPI-only coordination substrate for a pool of AI agents draining a backlog of bite-size tasks