惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Application and Cybersecurity Blog
Application and Cybersecurity Blog
A
About on SuperTechFans
S
SegmentFault 最新的问题
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
Help Net Security
Help Net Security
有赞技术团队
有赞技术团队
博客园 - 【当耐特】
O
OpenAI News
美团技术团队
月光博客
月光博客
Apple Machine Learning Research
Apple Machine Learning Research
Schneier on Security
Schneier on Security
Webroot Blog
Webroot Blog
Cyberwarzone
Cyberwarzone
Hacker News - Newest:
Hacker News - Newest: "LLM"
Google Online Security Blog
Google Online Security Blog
T
Tenable Blog
S
Security Affairs
博客园_首页
S
Schneier on Security
Security Latest
Security Latest
T
Threat Research - Cisco Blogs
T
Tailwind CSS Blog
大猫的无限游戏
大猫的无限游戏
Spread Privacy
Spread Privacy
量子位
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
K
Kaspersky official blog
Hugging Face - Blog
Hugging Face - Blog
TaoSecurity Blog
TaoSecurity Blog
博客园 - 聂微东
Vercel News
Vercel News
M
MIT News - Artificial intelligence
T
Troy Hunt's Blog
B
Blog
MongoDB | Blog
MongoDB | Blog
Martin Fowler
Martin Fowler
Attack and Defense Labs
Attack and Defense Labs
L
LINUX DO - 最新话题
D
DataBreaches.Net
cs.CV updates on arXiv.org
cs.CV updates on arXiv.org
Stack Overflow Blog
Stack Overflow Blog
cs.AI updates on arXiv.org
cs.AI updates on arXiv.org
博客园 - Franky
W
WeLiveSecurity
Threat Intelligence Blog | Flashpoint
Threat Intelligence Blog | Flashpoint
F
Fortinet All Blogs
www.infosecurity-magazine.com
www.infosecurity-magazine.com
C
Check Point Blog
H
Hacker News: Front Page

Hacker News: Show HN

PurrrrrFocus: Pomodoro Timer App - App Store Workflow Engine — Multi-Step Orchestration for Bun RapidPhoto: Pro Photo Editor App - App Store GitHub - DheerG/swarms: Achieve extraordinary results with claude code across a variety of tasks SPICE simulation → oscilloscope → verification with Claude Code — Lucas Gerads Show HN: VCoding – A 5 MB native Windows IDE with no dynamic dependencies Show HN: LLMs don't hallucinate because they're bad at math, it's the format GitHub - Agent-FM/agentfm-core: AgentFM is a peer-to-peer network that turns everyday computers into a decentralized AI supercomputer. AgentFM lets you run massive AI workloads directly across a global mesh of idle CPUs and GPUs. Show HN: Tracking Top US Science Olympiad Alumni over Last 25 Years GitHub - Potarix/agent-hub: One place to talk to all your agents Show HN: Runtime security for AI agents(injection,tool abuse, data exfiltration) GitHub - dubeyKartikay/lazyspotify: Terminal Spotify client for macOS and Linux GitHub - the-banana-tool/king-louie: Easy to use GUI Personal AI Assistant. Win/Linux/Mac. Show HN I made my vacation rental bookable by AI agents–no Airbnb, 0% commission GitHub - basteez/jsf-autoreload: maven plugin to enable hot reload on jsf projects uvm32/hosts/host-gdbstub at main · ringtailsoftware/uvm32 GitHub - labsai/EDDI: Config-driven engine that turns JSON into production-grade AI agents. Multi-agent orchestration, 12+ LLM providers, MCP/A2A protocols, RAG, persistent memory, and enterprise compliance (EU AI Act, GDPR, HIPAA). Built on Quarkus. GitHub - glitchnsec/fortyone-oss: AI Executive Assistant Platform Quickstart | Alien GitHub - muxshed/shed: One stream in, or many. Every destination, simultaneously. No cloud middleman, no per-channel fees, no limits. GitHub - ocrbase-hq/ocrbase: 📄 PDF/IMG ->.MD/JSON Document OCR API for PaddleOCR and GLMOCR. Self-hostable. GitHub - impactjo/home-memory: MCP server that lets your AI assistant remember everything about your home. GitHub - Sets88/dbcls: DbCls is a powerful terminal database client that supports various databases GitHub - neptun2000/heor-agent-mcp GitHub - SeanFDZ/macmind: Single-layer transformer in HyperTalk for the classic Macintosh RollQuation: Math Puzzles - Apps on Google Play GitHub - dropbox/witchcraft Show HN: Agent-cache – Multi-tier LLM/tool/session caching for Valkey and Redis GitHub - opentalon/opentalon: OpenTalon is an open-source platform built from the ground up in Go as a robust alternative to OpenClaw LinkedIn™ 职位抓取工具 - Chrome 应用商店 GitHub - EdoardoBambini/Agent-Armor-Iaga: AI agents are getting tool access — shell, file system, databases, APIs, secrets. But **nobody is governing what they actually do with it**. Frameworks like LangChain, CrewAI, AutoGen, and Claude Code give agents the power to execute. Agent Armor gives you the power to control, audit, and approve every single action before it happens. HN Vibes — Week 15, Apr 7–13 2026 GitHub - chojs23/ec: Easy terminal-native 3-way git mergetool vim-like workflow GitHub - SethPyle376/hiraeth: Local AWS emulator focused on fast integration testing, with SQS support, SQLite-backed state, and a debug-friendly web UI. GitHub - JakOb-dotcom/cloud-sandbox-security-analysis: Technical analysis and Proof of Concept (PoC) regarding environment variable exfiltration in containerized cloud sandboxes via side-channel data leaks. Springboards - Flint Alpha Show HN: A simpler coding agent harness GitHub - audiodude/sudomake-friends GitHub - 256thFission/mini-mythos: OSS clone of Anthropic’s Mythos harness to locate C/C++ memory vulnerabilities Show HN: OpenParallax: OS-level privilege separation for AI agent execution Hacker News Sorted - Chrome 应用商店 Show HN: How to Install Docker on Ubuntu 24.04 LTS: Complete 2026 Guide GitHub - himanshudongre/smriti GitHub - sverrirsig/claude-control: macOS desktop dashboard for monitoring and managing multiple Claude Code sessions GitHub - ory/dockertest: Write better integration tests! Dockertest helps you boot up ephermal docker images for your Go tests with minimal work. Chiral - Chrome 应用商店 Show HN: Two Claudes collaborating through shared memory on a $100 mini-PC GitHub - pmichaillat/latex-cv: Minimalist LaTeX template for academic CVs GitHub - oguzbilgic/posse: A web UI for Anthropic Managed Agents. GitHub - sshiraz/depsly: Dependency risk analysis tool for npm packages ABI Add safari/agent-harness — Safari browser automation via safari-mcp by achiya-automation · Pull Request #212 · HKUDS/CLI-Anything GitHub - Halfblood-Prince/trustcheck: Verify PyPI package attestations and improve Python supply-chain security GitHub - oguzbilgic/kern-ai: Agents that do the work and show it. GitHub - bruits/satteri: High-performance Markdown and MDX processing for the JavaScript ecosystem GitHub - tylergibbs1/feedstock: High-performance web crawler and scraper for TypeScript, powered by Bun and Playwright GitHub - Grimm67123/grimmbot: The self-improving sandboxed and open-source AI agent. With persistent memory and scheduling. GitHub - whitevanillaskies/whitebloom: Local whiteboard that blooms. GitHub - hwdsl2/docker-whisper: Docker image for a self-hosted Whisper speech-to-text server with speaker diarization and OpenAI-compatible transcription and translation APIs. Powered by faster-whisper. Supports all Whisper models, NVIDIA GPU (CUDA) acceleration, JSON/SRT/VTT output, SSE streaming, offline mode, and multi-arch (amd64, arm64). GitHub - yisding/reviewwiggum GitHub - MarwanAlsoltany/serrors: Structured errors for Go: sentinel hierarchies, typed data, custom formatting, and slog integration. GitHub - soatok/age-php GitHub - Luthiraa/markitme GitHub - stagas/rtdiff: realtime git diff gui and AI-assisted commits GitHub - tombedor/excalicharts GitHub - wh1le/excalidraw-edit: Open and edit .excalidraw files from the terminal. Offline, auto-saves to disk. MalExt Sentry - Malicious Extension Scanner - Chrome 应用商店 GitHub - syi0808/asciianimesvg: Generate animated ASCII art SVGs from text. CLI, Rust library, WASM, and web editor. GitHub - zaina-ml/ml_forge: A visual-based graph node editor for training computer vision models. GitHub - anakin87/llm-rl-environments-lil-course: 🌱 A little course on Reinforcement Learning Environments for evaluating and training Language Models GitHub - takaakit/superpowers-uml: Superpowers-UML modifies Superpowers to ensure a software development workflow in which AI agents design through UML modeling. AdriByte Studio - Sviluppo Web e Soluzioni Digitali GitHub - chouligi/angel-copilot: Your personalized Angel Investment Advisor Show HN: MoodSense AI (ML and FastAPI and Gradio, Deployed on Hugging Face) Moodsense Ai - a Hugging Face Space by aman179102 GitHub - agenteractai/lodmem: Level Of Detail Context Management for Agents GitHub - ostefani/subnetlens: A fast, concurrent network scanner with a TUI and plain-text CLI, built in Go. It discovers live hosts on your network, scans their open ports, resolves hostnames, and fingerprints operating systems—delivered. Cyber Pulse: Agentic Intel - Apps on Google Play Whisper API: Self-Hostable Speech to Text Transcription The Agent-Web Protocol Stack: A Research Thesis GitHub - msmarkgu/RelayFreeLLM: A restful API designed to route user prompts to various AI model providers. Show HN: Provepy – A Python decorator that proves your code using Lean and LLMs Show HN: Pardonned.com – A searchable database of US Pardons GitHub - patrickdappollonio/dux: Dux is a terminal UI that lets you run multiple AI coding agents side by side, each in its own git worktree, with full companion terminals, macros, commit generation, and a command palette that knows more tricks than you do. kMC Crystal Simulator Show HN: HyperFlow – A self-improving agent framework built on LangGraph GitHub - stef41/vibescore: 🎵 Grade your vibe-coded project. One command, instant letter grade across security, quality, dependencies, and testing. GitHub - stef41/lmscan: 🔍 Detect AI-generated text and fingerprint which LLM wrote it. Open-source GPTZero alternative. Zero dependencies, works offline. imgur.com GitHub - visionscaper/collabmem: Enabling long-term collaboration with Agentic AI - building up episodic and world model memory over time with in-context awareness 在 Steam 上购买 FriedrichAI: Offline AI 立省 10% GitHub - atripati/ark: AI Runtime Kernel — a context operating system for AI agents. Eliminates tool bloat, loads only what’s needed, and gives LLMs their reasoning space back. GitHub - nowork-studio/toprank: Open-source Claude Code skills for SEO, SEM, Google Ads GitHub - tacomanator/sash: Lightweight macOS menu bar app for reliably cycling through windows of the current application. Appents | Social Media Management for Product-First Teams GitHub - pnhoang/youtube-spam-blocker: Automatically detects and hides spam messages in YouTube Live chat. Set rate limits, keyword filters, and block repeat offenders. GitHub - decisionnode/DecisionNode: CLI + Local MCP - A shared structured memory store across Claude Code, Cursor, Windsurf, Antigravity, and every MCP client. Semantically queryable. GitHub - AvaCodeSolutions/django-email-learning: An open source Django app for creating email-based learning platforms with IMAP integration and React frontend components. The $100K Gap in Kubernetes Security Tooling Function Calling Harness: From 6.75% to 100%
GitHub - sofumel/claude-handoff-revive: Resume Claude Code work after rate/usage/context limits without replaying the prior transcript. Auto-saves at 90%/95% usage. Plugin-installable, 10 languages.
sofumel · 2026-05-27 · via Hacker News: Show HN

claude-handoff-revive

CI Release Stars Contributors License: MIT Multi-language

Continue Claude Code work after a rate / usage / context limit without replaying your prior session — typically tens of thousands of tokens, often 100k+ for long sessions.

🇺🇸 English · 🇯🇵 日本語 · 🇨🇳 中文 · 🇹🇼 繁體中文 · 🇰🇷 한국어 · 🇪🇸 Español · 🇵🇹 Português · 🇩🇪 Deutsch · 🇫🇷 Français · 🇹🇷 Türkçe


🇺🇸 English

The problem

When Claude Code shows You've hit your limit · resets ..., the standard recovery — claude --resume or claude -c — reloads the entire prior conversation into context. A medium session typically burns tens of thousands of tokens, often 100k+ for long sessions, before you've asked a single question.

What this skill does

It saves only the minimum needed to continue, structured into .claude/handoff/current.md (~1–3k tokens):

  • goal — what you were trying to accomplish
  • done / wip / todo — completed, in-progress, not-yet-started tasks
  • next_action — the concrete next step (file:line + exact command)
  • touched_files — what files were touched and why
  • decisions — design choices, with the reason behind each

To resume, start a new session (don't use --resume). The skill reads only that one file and picks up where you left off.

Install

Plugin install (recommended) — run these inside Claude Code:

/plugin marketplace add sofumel/claude-handoff-revive       # 1. register marketplace
/plugin install handoff-revive@handoff-revive-marketplace   # 2. install plugin

Skill, slash commands, and hooks auto-activate. No settings.json editing required.

/plugin isn't available in this environment? Your Claude Code version is too old. Check with claude --version and update via brew upgrade claude-code (Homebrew) or npm update -g @anthropic-ai/claude-code (npm). If you can't update, use the manual install below.

Manual install (always works, no plugin support required):

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

# Or globally for all projects:
./install.sh --global       # .\install.ps1 -Global

For the manual install, optionally enable hooks by merging the snippet from HOOK_SETUP.md into your .claude/settings.json (the plugin install does this automatically).

How to use

Quick reference:

Action Type this in Claude Code
Save handoff /handoff-revive:save
Save with extra quality check /handoff-revive:save-strict
Resume in a NEW session /handoff-revive:resume
Toggle auto-save for this session /handoff-revive:auto on / off / status

Step 1 — when you're approaching a rate limit, save

In Claude Code, type the following command:

/handoff-revive:save

Behind the scenes, the skill runs these steps automatically (all zero-LLM-token where possible):

  1. Detects your language (10 supported: en / ja / zh / zh-TW / ko / es / pt / de / fr / tr) from your message
  2. Pulls changed files from git status — you don't have to remember which files you edited
  3. Writes a schema-conformant handoff to .claude/handoff/current.md
  4. Validates the schema, cleans up dead paths, strips empty sections — pure shell, no tokens
  5. Displays the estimated savings in the chat

Step 2 — wait for the rate limit to reset

When the time on resets ... passes, your usage window opens again.

Step 3 — start a NEW session (do NOT use --resume)

claude

When the new session starts:

  • The SessionStart hook detects .claude/handoff/current.md (if modified within the last 7 days)
  • It injects context for Claude: "A recent handoff exists — prefer it over --resume"
  • You run /handoff-revive:resume and Claude reads only the handoff file and continues

Step 4 — (optional) higher-quality saves

/handoff-revive:save-strict

After the regular save, Claude runs an extra semantic-quality check:

  • Is goal specific enough that someone with no context understands it?
  • Does each decision include a real reason, not just the decision restated?
  • Are blockers phrased as answerable questions?
  • Does wip describe state precisely enough to resume mid-edit?

Vague fields get rewritten automatically.

Optional: turn-count nudge

If the Stop hook is enabled, Claude reminds you to checkpoint every N turns (default 15):

[handoff-revive] Turn 15 — checkpoint due. Run /handoff-revive:save to save.

Plugin install activates this automatically; manual install — see HOOK_SETUP.md. Override the threshold with HANDOFF_CHECKPOINT_EVERY=10.

Auto-save at 90% / 95% usage (synced with the UI notification)

The usage-monitor hook (PostToolUse) reads rate_limits.five_hour.used_percentage — the same value that drives Claude Code's "approaching usage limit" notification. When you cross 90%, Claude auto-saves a full handoff before its next response (no asking, no interruption). At 95%, it saves again with an urgent notice.

Per-session opt-out if you don't want auto-save in the current thread:

/handoff-revive:auto off       # disable for THIS session
/handoff-revive:auto on        # re-enable
/handoff-revive:auto status    # show current state + thresholds

New Claude sessions reset to enabled. To disable globally: export HANDOFF_AUTO_SAVE_PERCENT=disabled HANDOFF_URGENT_PERCENT=disabled.

To shift thresholds: export HANDOFF_AUTO_SAVE_PERCENT=80 (earlier), or set per-hook in settings.json (see HOOK_SETUP.md).

Token budget

Method Tokens replayed to resume
claude --resume tens of thousands, often 100k+
claude -c tens of thousands, often 100k+
handoff-revive 1,000–3,000

You break even after just one save and one resume. Even in the worst case where you save and never resume, the overhead is only a few KB — costs never run away.

License

MIT — see LICENSE.


🇯🇵 日本語

解決する問題

You've hit your limit · resets ... という制限通知が表示された後、claude --resume-c で再開すると、それまでの会話履歴がまるごとコンテキストに再ロードされます。中規模のセッションでも数万トークン、長いセッションでは 10 万を超えるトークン分が、まだ何も質問していない段階で消費されてしまいます。

この skill が提供するもの

作業の続きに必要な最小限の情報だけを構造化して .claude/handoff/current.md に保存します(約 1〜3k トークン)。保存される項目:

  • goal — 何を達成しようとしていたか
  • done / wip / todo — 完了済み / 進行中 / 未着手のタスク
  • next_action — 実行可能な次の一手(file:line + 具体的なコマンド)
  • touched_files — 触ったファイルとその理由
  • decisions — 設計判断と、その根拠

再開時は claude --resume を使わず 新規セッションを起動するだけ。skill がこの 1 ファイルだけを読み込み、即座に作業を引き継ぎます。

インストール

プラグインインストール (推奨) — Claude Code 内で実行:

/plugin marketplace add sofumel/claude-handoff-revive       # 1. マーケットプレイス登録
/plugin install handoff-revive@handoff-revive-marketplace   # 2. プラグインインストール

Skill・スラッシュコマンド・hook がすべて自動で有効化されます。settings.json の編集は不要。

/plugin isn't available in this environment と表示された場合、Claude Code のバージョンが古い可能性があります。claude --version で確認し、brew upgrade claude-code (Homebrew) または npm update -g @anthropic-ai/claude-code (npm) で更新してください。それでも使えない場合は下のマニュアルインストールを使用してください。

マニュアルインストール (どの環境でも動作):

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

# 全プロジェクト共通でインストール:
./install.sh --global       # .\install.ps1 -Global

マニュアルインストールの場合のみ、HOOK_SETUP.md のスニペットを .claude/settings.json にマージして hooks を有効化してください (プラグインインストールでは自動)。

使い方

早見表:

やりたいこと Claude Code に入力するコマンド
ハンドオフを保存 /handoff-revive:save
品質チェック付きで保存 /handoff-revive:save-strict
新規セッション で再開 /handoff-revive:resume
このセッションの自動保存を切り替え /handoff-revive:auto on / off / status

ステップ 1 — レート制限が近づいたら保存する

Claude Code で、以下のコマンドを入力してください:

/handoff-revive:save

裏では skill が以下を自動実行します(できる限りトークンを使わずに):

  1. 言語を検出 (10 言語対応: en / ja / zh / zh-TW / ko / es / pt / de / fr / tr) — あなたのメッセージから判定
  2. git status から変更ファイルを抽出 — どのファイルを編集したか思い出す必要なし
  3. スキーマ通りのハンドオフを .claude/handoff/current.md に書き込み
  4. スキーマ検証 + 無効パスエントリの自動削除 + 空セクションの除去 — すべて純シェル (Claude を呼ばずシェルスクリプトだけで処理) なので LLM トークンを消費しません
  5. 節約見込みをチャットに表示

ステップ 2 — レート制限の解除を待つ

resets ... の時刻を過ぎたら、使用枠が回復します。

ステップ 3 — 新規セッションを起動 (--resume は使わない)

claude

新セッションが起動すると:

  • SessionStart hook が .claude/handoff/current.md (7 日以内に更新) を検出
  • Claude にコンテキスト注入: 「最近のハンドオフがあります。--resume ではなくこちらを優先してください」
  • あなたが /handoff-revive:resume を実行すると、Claude がハンドオフファイルだけを読んで作業再開

ステップ 4 — (任意) 保存品質を強化したい場合

/handoff-revive:save-strict

通常の保存後に、Claude による追加の意味的チェックが走ります:

  • goal が具体的かどうか (「認証バグを直す」のような曖昧さでなく「login.ts の bcrypt.compare をタイミング攻撃に強い実装に置換」のような具体性)
  • decision根拠が書いてあるか (決定の言い換えではなく "WHY" が明示されているか)
  • blockers が「答えられる質問形式」になっているか
  • wip が「再開時に作業位置が分かる粒度」になっているか

不十分なフィールドを Claude が自動的に書き直します。

任意機能: ターン数 nudge

Stop hook を有効化すると、ターン数が指定値(デフォルト 15)を超えるたびにリマインドが出ます:

[handoff-revive] Turn 15 — checkpoint due. Run /handoff-revive:save to save.

プラグインインストールでは自動有効化されます。マニュアルインストールの場合は HOOK_SETUP.md を参照。閾値は環境変数 HANDOFF_CHECKPOINT_EVERY=10 で変更可能。

90% / 95% 到達で自動保存 (UI 通知と同期)

usage-monitor hook (PostToolUse) が rate_limits.five_hour.used_percentage を読み取ります — Claude Code の「使用量の上限に近づいています」通知を駆動するのと同じ値です。90% 到達時、Claude が次の応答前にフルクオリティのハンドオフを自動保存します (確認なし、作業中断なし)。95% に達すると、緊急通知付きで再度自動保存されます。

このセッションの自動保存をオフにしたい場合:

/handoff-revive:auto off       # このセッションのみ無効化
/handoff-revive:auto on        # 再有効化
/handoff-revive:auto status    # 現在の状態 + 閾値を表示

新しい Claude セッションを起動すると自動保存はデフォルト ON に戻ります。グローバルに無効化するには: export HANDOFF_AUTO_SAVE_PERCENT=disabled HANDOFF_URGENT_PERCENT=disabled

閾値変更: export HANDOFF_AUTO_SAVE_PERCENT=80 (より早めに発火)、または settings.json の hook で個別指定 (HOOK_SETUP.md 参照)。

トークン収支

方式 再開時に再ロードされるトークン
claude --resume 数万、長セッションでは 10 万超も
claude -c 数万、長セッションでは 10 万超も
handoff-revive 1,000〜3,000

1 回保存して 1 回再開すれば、必ず黒字です。 たとえ保存だけで一度も再開しない場合でも、発生するのは数 KB 程度のオーバーヘッドだけで、コストが暴走することはありません。

ライセンス

MIT — LICENSE 参照。


🇨🇳 中文

要解决的问题

当 Claude Code 显示 You've hit your limit · resets ... 时,使用 claude --resumeclaude -c 恢复会把之前的整个会话重新载入上下文。中等规模会话通常需要数万 tokens,长会话经常超过 10 万——还没问一个问题就已经用掉了。

此 skill 提供什么

把恢复所需的最少信息以结构化方式保存到 .claude/handoff/current.md(约 1–3k tokens):

  • goal — 你想完成什么
  • done / wip / todo — 完成 / 进行中 / 待办
  • next_action — 可执行的下一步(file:line + 具体命令)
  • touched_files — 涉及的文件及其原因
  • decisions — 设计决策及其理由

恢复时启动新会话(不要用 --resume)。skill 只读取这一个文件即可继续工作。

安装

插件安装 (推荐) — 在 Claude Code 内运行:

/plugin marketplace add sofumel/claude-handoff-revive       # 1. 注册插件市场
/plugin install handoff-revive@handoff-revive-marketplace   # 2. 安装插件

Skill、斜杠命令、hooks 全部自动激活,无需编辑 settings.json

如果出现 /plugin isn't available in this environment,说明 Claude Code 版本过旧。运行 claude --version 确认,并通过 brew upgrade claude-codenpm update -g @anthropic-ai/claude-code 升级。

手动安装 (始终可用):

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

手动安装时可选:将 HOOK_SETUP.md 中的代码片段合并到 .claude/settings.json 启用 hooks (插件安装会自动完成)。

使用方法

快速参考:

操作 在 Claude Code 输入
保存 handoff /handoff-revive:save
带质量检查的保存 /handoff-revive:save-strict
新会话中恢复 /handoff-revive:resume
切换本会话的自动保存 /handoff-revive:auto on / off / status

步骤 1 — 接近速率限制时保存

/handoff-revive:save

skill 自动执行(尽量零 token):

  1. 检测语言 (10 种语言: en / ja / zh / zh-TW / ko / es / pt / de / fr / tr)
  2. git status 提取变更文件 — 不用回忆改了哪些文件
  3. 按 schema 写入 .claude/handoff/current.md
  4. schema 验证 + 死路径删除 + 空段落删除 — 纯 shell,零 token
  5. 在聊天中显示节省估算

步骤 2 — 等待限制重置

到达 resets ... 时间后,使用额度恢复。

步骤 3 — 启动新会话 (不要用 --resume)

claude

新会话启动时:

  • SessionStart hook 检测 .claude/handoff/current.md (7 天以内)
  • 注入上下文: "存在最近的 handoff,优先使用而不是 --resume"
  • 你运行 /handoff-revive:resume — Claude 只读 handoff 文件即可恢复

步骤 4 — (可选) 提高保存质量

/handoff-revive:save-strict

常规保存后再追加 Claude 的语义检查:

  • goal 是否具体?
  • 各 decision 是否包含真实理由?
  • blockers 是否为可回答的问题?
  • wip 是否精确到能从中间恢复?

90% / 95% 自动保存 (与 UI 通知同步)

usage-monitor hook 读取 rate_limits.five_hour.used_percentage —— 这与 Claude Code "接近使用上限" 通知所用的值相同。90% 时 Claude 在下次回复前自动完整保存 handoff (无需询问、不打断)。95% 时再次保存并附紧急通知。

本会话不需要自动保存时:

/handoff-revive:auto off       # 仅本会话禁用
/handoff-revive:auto on        # 重新启用
/handoff-revive:auto status    # 查看状态 + 阈值

新会话默认启用。全局禁用: export HANDOFF_AUTO_SAVE_PERCENT=disabled HANDOFF_URGENT_PERCENT=disabled

License

MIT — see LICENSE.


🇹🇼 繁體中文

要解決的問題

當 Claude Code 顯示 You've hit your limit · resets ... 時,使用 claude --resumeclaude -c 恢復會把之前的整個會話重新載入上下文。中等規模對話通常需要數萬 tokens,長對話經常超過 10 萬——還沒問一個問題就已經用掉了。

此 skill 提供什麼

把恢復所需的最少資訊以結構化方式儲存到 .claude/handoff/current.md(約 1–3k tokens):

  • goal — 你想完成什麼
  • done / wip / todo — 完成 / 進行中 / 待辦
  • next_action — 可執行的下一步(file:line + 具體命令)
  • touched_files — 涉及的檔案及其原因
  • decisions — 設計決策及其理由

恢復時啟動新會話(不要用 --resume)。skill 只讀取這一個檔案即可繼續工作。

安裝

外掛程式安裝 (推薦) — 在 Claude Code 內執行:

/plugin marketplace add sofumel/claude-handoff-revive       # 1. 註冊 marketplace
/plugin install handoff-revive@handoff-revive-marketplace   # 2. 安裝外掛

如果出現 /plugin isn't available in this environment,說明 Claude Code 版本過舊。執行 claude --version 確認,並透過 brew upgrade claude-codenpm update -g @anthropic-ai/claude-code 升級。

手動安裝 (始終可用):

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

使用方式

操作 在 Claude Code 輸入
儲存 handoff /handoff-revive:save
帶品質檢查的儲存 /handoff-revive:save-strict
新會話中恢復 /handoff-revive:resume
切換本會話的自動儲存 /handoff-revive:auto on / off / status

License

MIT — see LICENSE.


🇰🇷 한국어

해결하는 문제

You've hit your limit · resets ... 메시지가 나타난 뒤 claude --resume 또는 -c 로 재개하면 이전 세션 전체 가 컨텍스트로 다시 로드됩니다. 중간 규모 세션이라면 수만 토큰, 긴 세션에서는 10만+ 토큰이 아무 질문하기도 전에 사라집니다.

이 skill 이 제공하는 것

작업 재개에 필요한 최소 정보만을 구조화하여 .claude/handoff/current.md 에 저장합니다(약 1–3k tokens):

  • goal — 무엇을 달성하려 했는가
  • done / wip / todo — 완료 / 진행 중 / 대기
  • next_action — 실행 가능한 다음 단계 (file:line + 구체적 명령)
  • touched_files — 다룬 파일과 그 이유
  • decisions — 설계 결정과 근거

재개할 때는 --resume 을 쓰지 말고 새 세션을 시작합니다. skill 이 그 파일 하나만 읽고 즉시 작업을 이어갑니다.

설치

플러그인 설치 (권장) — Claude Code 안에서 실행:

/plugin marketplace add sofumel/claude-handoff-revive       # 1. 마켓플레이스 등록
/plugin install handoff-revive@handoff-revive-marketplace   # 2. 플러그인 설치

/plugin isn't available in this environment 메시지가 나오면 Claude Code 버전이 오래되었습니다. claude --version 으로 확인하고 brew upgrade claude-code 또는 npm update -g @anthropic-ai/claude-code 로 업데이트하세요.

수동 설치 (항상 동작):

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

사용법

빠른 참조:

작업 Claude Code 입력
핸드오프 저장 /handoff-revive:save
품질 검사 포함 저장 /handoff-revive:save-strict
새 세션 에서 재개 /handoff-revive:resume
이 세션 자동 저장 토글 /handoff-revive:auto on / off / status

단계 1 — 사용량 한도가 가까워지면 저장

/handoff-revive:save

skill 이 자동으로 (가능한 한 0 토큰으로):

  1. 언어 감지 (10개 언어 지원: en / ja / zh / zh-TW / ko / es / pt / de / fr / tr)
  2. git status 에서 변경 파일 추출
  3. 스키마대로 .claude/handoff/current.md 에 저장
  4. 스키마 검증 + 죽은 경로 제거 + 빈 섹션 제거 — 순 shell, 0 토큰
  5. 절약량 표시

단계 2 — 한도 리셋 대기

resets ... 시간이 지나면 사용량이 복구됩니다.

단계 3 — 새 세션 시작 (--resume 사용 금지)

claude

새 세션 시작 시:

  • SessionStart hook 이 .claude/handoff/current.md 감지 (7일 이내)
  • 컨텍스트 주입: "최근 핸드오프 존재 — --resume 대신 이걸 사용"
  • /handoff-revive:resume 를 실행하면 Claude 가 핸드오프만 읽고 재개

단계 4 — (선택) 더 높은 품질의 저장

/handoff-revive:save-strict

추가로 Claude 의 의미적 검증 실행: goal 이 구체적인지, 각 decision 에 진짜 이유가 있는지 등.

90% / 95% 자동 저장 (UI 알림과 동기)

usage-monitor hook 이 rate_limits.five_hour.used_percentage 를 읽음 — Claude Code 의 "사용량 한도 접근" 알림과 같은 값. 90% 도달 시 Claude 가 다음 응답 전에 풀 핸드오프를 자동 저장 (질문 없이, 중단 없이). 95% 에서 긴급 알림과 함께 재저장.

이 세션에서 자동 저장 끄기:

/handoff-revive:auto off       # 이 세션만 비활성화
/handoff-revive:auto on        # 다시 활성화
/handoff-revive:auto status    # 상태 + 임계값 확인

새 세션은 기본 활성화. 전역 비활성화: export HANDOFF_AUTO_SAVE_PERCENT=disabled HANDOFF_URGENT_PERCENT=disabled.

License

MIT — see LICENSE.


🇪🇸 Español

El problema

Cuando Claude Code muestra You've hit your limit · resets ..., la recuperación habitual — claude --resume o -c — recarga toda la conversación anterior en el contexto. Una sesión mediana consume decenas de miles de tokens, a menudo 100k+ en sesiones largas, antes de hacer una sola pregunta.

Lo que ofrece este skill

Guarda solo el mínimo necesario para continuar, estructurado en .claude/handoff/current.md (~1–3k tokens):

  • goal — qué intentabas lograr
  • done / wip / todo — completado / en progreso / pendiente
  • next_action — siguiente paso ejecutable (file:line + comando exacto)
  • touched_files — archivos tocados y su razón
  • decisions — decisiones de diseño y el por qué

Para reanudar, inicia una sesión nueva (NO uses --resume). El skill lee solo ese archivo y continúa.

Instalación

Plugin (recomendado) — ejecutar dentro de Claude Code:

/plugin marketplace add sofumel/claude-handoff-revive       # 1. registrar marketplace
/plugin install handoff-revive@handoff-revive-marketplace   # 2. instalar plugin

Si aparece /plugin isn't available in this environment, tu versión de Claude Code es antigua. Verifica con claude --version y actualiza con brew upgrade claude-code o npm update -g @anthropic-ai/claude-code.

Instalación manual (siempre funciona):

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

Cómo usar

Referencia rápida:

Acción Escribe en Claude Code
Guardar handoff /handoff-revive:save
Guardar con verificación /handoff-revive:save-strict
Reanudar en una sesión nueva /handoff-revive:resume
Alternar auto-guardado en esta sesión /handoff-revive:auto on / off / status

Paso 1 — al acercarte al límite, guarda

/handoff-revive:save

El skill ejecuta automáticamente (con cero tokens cuando es posible):

  1. Detecta tu idioma (10 idiomas: en / ja / zh / zh-TW / ko / es / pt / de / fr / tr)
  2. Extrae archivos modificados de git status
  3. Escribe el handoff conforme al schema en .claude/handoff/current.md
  4. Valida schema + elimina rutas muertas + elimina secciones vacías — solo shell, cero tokens
  5. Muestra el ahorro estimado

Paso 2 — espera al reset

Cuando pase la hora indicada en resets ..., tu cuota se restablece.

Paso 3 — inicia una sesión nueva (NO uses --resume)

claude

Al iniciar:

  • El hook SessionStart detecta .claude/handoff/current.md (modificado en últimos 7 días)
  • Inyecta contexto: "Existe handoff reciente — prefiérelo sobre --resume"
  • Ejecuta /handoff-revive:resume — Claude lee solo el handoff y continúa

Paso 4 — (opcional) guardado de mayor calidad

/handoff-revive:save-strict

Tras el guardado normal, Claude ejecuta una verificación semántica: ¿es goal específico? ¿incluye cada decision una razón real?

Auto-guardado al 90% / 95% (sincronizado con la notificación de UI)

El hook usage-monitor lee rate_limits.five_hour.used_percentage — el mismo valor que dispara la notificación "uso cerca del límite" de Claude Code. Al 90% Claude auto-guarda un handoff completo antes de su siguiente respuesta (sin preguntar, sin interrumpir). Al 95% vuelve a guardar con aviso urgente.

Para desactivar el auto-guardado en esta sesión:

/handoff-revive:auto off       # desactivar SOLO esta sesión
/handoff-revive:auto on        # reactivar
/handoff-revive:auto status    # ver estado + umbrales

Las sesiones nuevas vuelven al estado activado. Desactivación global: export HANDOFF_AUTO_SAVE_PERCENT=disabled HANDOFF_URGENT_PERCENT=disabled.

License

MIT — see LICENSE.


🇵🇹 Português

O problema

Quando o Claude Code mostra You've hit your limit · resets ..., a recuperação habitual — claude --resume ou -c — recarrega toda a conversa anterior no contexto. Uma sessão média gasta dezenas de milhares de tokens, frequentemente 100k+ em sessões longas, antes de fazer uma única pergunta.

O que esta skill faz

Guarda apenas o mínimo necessário para continuar, estruturado em .claude/handoff/current.md (~1–3k tokens):

  • goal — o que estavas a tentar alcançar
  • done / wip / todo — concluído / em curso / pendente
  • next_action — próximo passo executável (file:line + comando exato)
  • touched_files — ficheiros tocados e a razão
  • decisions — decisões de design e o porquê

Para retomar, inicia uma sessão nova (NÃO uses --resume). A skill lê apenas esse ficheiro e continua.

Instalação

Plugin (recomendado) — executar dentro do Claude Code:

/plugin marketplace add sofumel/claude-handoff-revive       # 1. registar marketplace
/plugin install handoff-revive@handoff-revive-marketplace   # 2. instalar plugin

Se aparecer /plugin isn't available in this environment, a tua versão do Claude Code é antiga. Verifica com claude --version e atualiza via brew upgrade claude-code ou npm update -g @anthropic-ai/claude-code.

Instalação manual (sempre funciona):

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

Uso

Ação Escreve no Claude Code
Guardar handoff /handoff-revive:save
Guardar com verificação /handoff-revive:save-strict
Retomar numa sessão nova /handoff-revive:resume
Alternar auto-save nesta sessão /handoff-revive:auto on / off / status

License

MIT — see LICENSE.


🇩🇪 Deutsch

Das Problem

Wenn Claude Code You've hit your limit · resets ... anzeigt, lädt die übliche Wiederherstellung — claude --resume oder -cdie gesamte vorherige Konversation wieder in den Kontext. Eine mittelgroße Sitzung verbraucht zehntausende Tokens, oft 100k+ bei langen Sitzungen, bevor du eine einzige Frage gestellt hast.

Was dieser Skill tut

Speichert nur das Minimum, das zum Fortsetzen nötig ist, strukturiert in .claude/handoff/current.md (~1–3k Tokens):

  • goal — was du erreichen wolltest
  • done / wip / todo — erledigt / in Bearbeitung / ausstehend
  • next_action — konkreter nächster Schritt (file:line + exaktes Kommando)
  • touched_files — bearbeitete Dateien und der Grund
  • decisions — Design-Entscheidungen und das Warum

Zum Fortsetzen startest du eine neue Sitzung (KEIN --resume). Der Skill liest nur diese eine Datei und setzt fort.

Installation

Plugin (empfohlen) — in Claude Code ausführen:

/plugin marketplace add sofumel/claude-handoff-revive       # 1. Marketplace registrieren
/plugin install handoff-revive@handoff-revive-marketplace   # 2. Plugin installieren

Falls /plugin isn't available in this environment erscheint, ist deine Claude Code Version zu alt. Prüfe mit claude --version und aktualisiere via brew upgrade claude-code oder npm update -g @anthropic-ai/claude-code.

Manuelle Installation (funktioniert immer):

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

Benutzung

Aktion In Claude Code eingeben
Handoff speichern /handoff-revive:save
Mit Qualitätsprüfung speichern /handoff-revive:save-strict
In neuer Sitzung fortsetzen /handoff-revive:resume
Auto-Save für diese Sitzung umschalten /handoff-revive:auto on / off / status

License

MIT — see LICENSE.


🇫🇷 Français

Le problème

Quand Claude Code affiche You've hit your limit · resets ..., la récupération habituelle — claude --resume ou -c — recharge toute la conversation précédente dans le contexte. Une session de taille moyenne consomme des dizaines de milliers de tokens, souvent 100k+ pour les sessions longues, avant que vous n'ayez posé une seule question.

Ce que fait ce skill

Sauvegarde uniquement le minimum nécessaire pour continuer, structuré dans .claude/handoff/current.md (~1–3k tokens) :

  • goal — ce que vous essayiez d'accomplir
  • done / wip / todo — terminé / en cours / à faire
  • next_action — prochaine étape exécutable (file:line + commande exacte)
  • touched_files — fichiers modifiés et la raison
  • decisions — décisions de conception et le pourquoi

Pour reprendre, lancez une nouvelle session (PAS --resume). Le skill lit uniquement ce fichier et continue.

Installation

Plugin (recommandé) — exécuter dans Claude Code :

/plugin marketplace add sofumel/claude-handoff-revive       # 1. enregistrer le marketplace
/plugin install handoff-revive@handoff-revive-marketplace   # 2. installer le plugin

Si /plugin isn't available in this environment apparaît, votre version de Claude Code est trop ancienne. Vérifiez avec claude --version et mettez à jour via brew upgrade claude-code ou npm update -g @anthropic-ai/claude-code.

Installation manuelle (fonctionne toujours) :

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

Utilisation

Action Tapez dans Claude Code
Sauvegarder le handoff /handoff-revive:save
Sauvegarder avec vérification qualité /handoff-revive:save-strict
Reprendre dans une NOUVELLE session /handoff-revive:resume
Basculer l'auto-sauvegarde pour cette session /handoff-revive:auto on / off / status

License

MIT — see LICENSE.


🇹🇷 Türkçe

Sorun

Claude Code You've hit your limit · resets ... mesajını gösterdiğinde, olağan kurtarma — claude --resume veya -ctüm önceki konuşmayı bağlama yeniden yükler. Orta ölçekli bir oturum, tek bir soru sormadan önce on binlerce token, uzun oturumlarda sıklıkla 100k+ tüketir.

Bu skill ne yapar

Devam etmek için yalnızca gerekli olan minimumu yapılandırılmış olarak .claude/handoff/current.md dosyasına kaydeder (~1–3k token):

  • goal — neyi başarmaya çalıştığınız
  • done / wip / todo — tamamlandı / devam ediyor / beklemede
  • next_action — somut sonraki adım (file:line + tam komut)
  • touched_files — dokunulan dosyalar ve nedeni
  • decisions — tasarım kararları ve nedenleri

Devam etmek için yeni bir oturum başlatın (--resume KULLANMAYIN). Skill yalnızca bu dosyayı okur ve devam eder.

Kurulum

Plugin (önerilen) — Claude Code içinde çalıştırın:

/plugin marketplace add sofumel/claude-handoff-revive       # 1. marketplace'i kaydet
/plugin install handoff-revive@handoff-revive-marketplace   # 2. plugin'i yükle

/plugin isn't available in this environment görürseniz, Claude Code sürümünüz çok eski. claude --version ile kontrol edin ve brew upgrade claude-code veya npm update -g @anthropic-ai/claude-code ile güncelleyin.

Manuel kurulum (her zaman çalışır):

git clone https://github.com/sofumel/claude-handoff-revive.git
cd claude-handoff-revive
./install.sh /path/to/your-project        # Linux/macOS/WSL/Git-Bash
# .\install.ps1 -Target C:\path\to\proj   # Windows PowerShell

Kullanım

Eylem Claude Code'a yazın
Handoff kaydet /handoff-revive:save
Kalite kontrolüyle kaydet /handoff-revive:save-strict
YENİ oturumda devam et /handoff-revive:resume
Bu oturum için otomatik kaydı değiştir /handoff-revive:auto on / off / status

License

MIT — see LICENSE.