By tobyilee
Session lifecycle management — write a forward-looking handoff before ending, verify-and-resume on the next session, and log conversations for long-running work. Includes opt-in pre-commit conversation-log enforcement hook.
Use this skill at the START of a new Claude Code session to pick up where the previous session left off. Finds the most recent handoff document under `.claude/reports/handoff/`, reads it, then VERIFIES its claims against the live codebase (Read each referenced file, run `git status` / `git log --oneline`) before reporting findings and waiting for user instruction. Treats the handoff as a hypothesis, not fact — previous session may have been confused. Trigger on "catchup", "catch up", "이어서 작업", "이어서 해줘", "이전 세션 이어", "resume from handoff", "handoff 읽어", "handoff 참고해서 시작", "어제 작업 이어서", "지난번 이어서", "새 세션 준비", "prepare to continue", "pick up where we left off", "캐치업". Also trigger when the user starts a new session by referencing prior work ("어제 한 작업 이어서 하자", "계속 하자", "resume the auth refactor") and an `.claude/reports/handoff/` directory exists in the project. Do NOT trigger when the user wants to start a brand-new unrelated task, or when they want to read a specific named document (use Read directly).
Use this skill to write a forward-looking handoff document that prepares the NEXT Claude Code session to pick up where this one stops. Generates a compact markdown file (~2K tokens) under `.claude/reports/handoff/` with Summary, Key Decisions, Traps to Avoid, Working Agreements, Relevant Files (with line numbers), Open Work (status-form only, never imperative), and a Prompt for New Chat that includes a verification instruction. Trigger on "handoff", "hand off", "세션 인계", "인계 문서", "다음 세션으로 넘겨", "end-of-session handoff", "session handoff", "세션 종료 전에 정리해", "handoff 문서 만들어", "handoff 써줘", "write a handoff", "prepare next session", "세션 넘기기 전에", "핸드오프". Also trigger when the user is about to run `/clear`, close the terminal, end the day, or says something like "let's wrap up this session", "세션 끝내기 전에 정리하자". Do NOT trigger for `save-conversation` (which is a retrospective transcript log, not a forward-looking handoff), git commit messages, PRD writing, or CLAUDE.md updates.
Use this skill to save a summary of the current conversation to a markdown file. Trigger on "save conv", "save conversation", "대화 저장", "대화내용 저장", "대화 저장해줘", "대화내용 저장해줘", "conversation 저장", "로그 저장", "대화 기록 남겨", "save chat", "save log", "대화 정리해서 저장", "지금까지 대화 저장", "conv 저장". Also trigger when the user asks to log, archive, or record the current conversation, or wants a summary of what was discussed saved to a file. Do NOT trigger for memory saving (use memory system) or for git commit messages.
Executes bash commands
Hook triggers when Bash tool is used
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Toby's personal Claude Code plugin marketplace.
v2.0.0 (2026-05-06) — The single
toby-essentialsplugin has been split into 5 focused plugins. See CHANGELOG.md for migration notes.
| Plugin | Version | What it's for |
|---|---|---|
toby-multi-agent | 1.0.0 | Delegate to Codex/Gemini, or run them as a persistent team in cmux |
toby-session | 1.0.0 | Session handoff/catchup, conversation logging |
toby-claude-config | 1.0.0 | Backup/restore ~/.claude, permissions merge, security hooks |
toby-codereview | 1.0.0 | /code-quality, /code-explore, PRD generation |
toby-personal | 1.0.0 | Toby-specific: Facebook style, Spring Boot, OMC tips, harness |
Multi-agent workspace. Delegate tasks to OpenAI Codex or Google Gemini, or spawn a persistent team in cmux panes alongside Claude. Default model versions live in plugins/toby-multi-agent/MODELS.md.
| Skill | Version | Description |
|---|---|---|
codex-delegate | 0.3.0 | Delegate tasks to OpenAI Codex CLI — background execution, context gathering, parallel delegation support |
gemini-delegate | 0.3.0 | Delegate tasks to Google Gemini CLI — background execution, context gathering, parallel delegation support |
cmux | 0.2.0 | Control the cmux terminal app — window/workspace/pane/surface management, browser automation, SSH, hooks integration |
toby-team-starter | 0.1.0 | Spawn Codex (--full-auto) and Gemini (--yolo) in cmux panes alongside Claude |
toby-codex | 0.1.0 | Send a task to the Codex cmux pane and collect the result as a file in tobyteam/ — requires "toby codex" prefix |
toby-gemini | 0.1.0 | Send a task to the Gemini cmux pane and collect the result as a file in tobyteam/ — requires "toby gemini" prefix |
tdd-team | 0.3.0 | 3-agent TDD team (Red/Green/Refactor) with task decomposition, progress tracking, user checkpoints |
Session lifecycle — handoff at end, catchup at start, conversation logs in between.
| Skill | Version | Description |
|---|---|---|
handoff | 0.1.0 | Write a forward-looking handoff under .claude/reports/handoff/handoff-<timestamp>.md before ending a session — Summary, Decisions, Traps, line-numbered file refs, status-form Open Work, ~2K token budget |
catchup | 0.1.0 | Resume from the latest handoff — Read-verifies every referenced file, cross-checks git status / git log, reports ✅Confirmed / ⚠️Shifted / ❌Missing, then stops for instruction |
save-conversation | — | Save a conversation summary to markdown, with Korean→English prompt rewrites logged to monthly prompt-{yyyymm}.md |
| Hook | Event | Description |
|---|---|---|
save-conv-before-commit.sh | PreToolUse (Bash) | Opt-in: only enforced in repos that already have conv-logs/ at their root. Blocks git commit unless a fresh conversation log is staged |
Manage and protect your global ~/.claude configuration.
| Command | Description |
|---|---|
/claude-backup | Create a timestamped backup of ~/.claude/ under ~/.claude-backup/claude-<YYYYMMDD-HHMMSS>/ |
/claude-backup-list | List existing ~/.claude-backup/ snapshots, newest first, with size + timestamp |
/claude-restore | Restore from a snapshot. Dry-run by default; --apply to overwrite, --backup PATH to target a specific snapshot |
/merge-permissions | Merge local project .claude/settings.local.json permissions into global ~/.claude/settings.json |
| Skill | Version | Description |
|---|---|---|
claude-backup | 0.1.0 | Back up and restore the user-authored parts of ~/.claude/. Snapshots under ~/.claude-backup/ (sibling of ~/.claude/, survives reinstall). Restore is dry-run by default and additive |
| Hook | Event | Description |
|---|---|---|
block-dangerous.sh | PreToolUse (Bash) | Block rm -rf outside project dir (direct and indirect via bash -c, eval). Uses realpath to resolve symlinks |
Parallel-agent code analysis and product spec writing.
| Command | Description |
|---|---|
/code-quality | Evaluate code quality with 4 parallel agents, score 9 dimensions (readability, maintainability, testability, performance, security, …), generate markdown report |
/code-explore | Deep codebase analysis with 5 parallel agents (structure, style, architecture, complexity, testing), generate comprehensive report |
npx claudepluginhub tobyilee/toby-plugins --plugin toby-sessionApplies Team Science's 'Big Five of Teamwork' (Salas, Sims & Burke 2005) to Claude Code agent teams. Ships 3 agents + 4 skills that enforce shared mental models, closed-loop communication, mutual performance monitoring, calibrated trust, and after-action debriefs so collaborative tasks — code, research, writing, creative — run as a high-performing team.
워크플로우 네이티브 메타 하네스 — 도메인을 위한 에이전트 팀·스킬·결정적 JavaScript Workflow 오케스트레이터(agent/pipeline/parallel/schema/budget/resume)를 한 번에 생성하는 빌더. v1 프로즈 하네스의 재개발판.
Session log archiver: saves conversations as Markdown or JSON for easy review
Mission Control for Claude Code — auto-starts a web dashboard, provides 90 MCP tools (sessions, stats, live monitoring, projects, teams, insights, coaching, and more), and adds 9 skills including /session-recap, /daily-cost, /standup, /coaching, /insights, and /team-status.
Provides an agent, skill, hook, and slash command related to the `aichat` tool-set for searching and resuming CLI-agent sessions
Loop engineering for AI coding agents — turn ad-hoc sessions into a repeatable research → plan → wave-execute → close loop with verification gates. Runs on Claude Code, Codex CLI, Cursor, and Pi.
Claude Code operations suite that bundles archive-aware local conversation discovery across Claude Code and Codex, internal-timestamp full-event session search and recovery, interrupted-work continuation, plugin/skill troubleshooting, CLAUDE.md progressive disclosure optimization, statusline configuration, exported .txt repair, full claude.ai conversation extraction with tool-call rendering and file download, plugin marketplace development, writing/testing/debugging Claude Code hooks, multi-provider profile isolation for running Kimi/GLM/DeepSeek/StepFun/Anthropic in separate windows, local source sync for Claude/Codex skill installs, personal-memory migration into tool-agnostic AGENTS.md reference docs, and terminal-output-to-PNG rendering for visual CLI verification under one shared namespace. Install once to get the full Claude Code power-user toolkit.
Ultra-compressed communication mode. Cuts 65% of output tokens (measured) while keeping full technical accuracy by speaking like a caveman.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claim