{
  "markdown": "# Plugins for Claude Natives\n\nA collection of Claude Code plugins for power users who want to extend Claude Code's capabilities beyond the defaults.\n\n## Table of Contents\n\n- [Quick Start](#quick-start)\n- [Available Plugins](#available-plugins)\n- [Plugin Details](#plugin-details)\n  - [agent-council](#agent-council) - Get consensus from multiple AI models\n  - [clarify](#clarify) - Transform vague requirements into specs\n  - [dev](#dev) - Community scanning + technical decision-making\n  - [doubt](#doubt) - Force Claude to re-validate responses\n  - [interactive-review](#interactive-review) - Review plans with a web UI\n  - [say-summary](#say-summary) - Hear responses via text-to-speech\n  - [youtube-digest](#youtube-digest) - Summarize and quiz on YouTube videos\n  - [gmail](#gmail) - Multi-account Gmail integration\n  - [google-calendar](#google-calendar) - Multi-account calendar integration\n  - [kakaotalk](#kakaotalk) - Send/read KakaoTalk messages on macOS\n  - [session-wrap](#session-wrap) - Session wrap-up + history analysis toolkit\n  - [team-assemble](#team-assemble) - Dynamic agent team orchestration\n  - [podcast](#podcast) - Source-to-YouTube Korean podcast generator\n  - [fetch-tweet](#fetch-tweet) - Fetch tweet text and metadata without auth\n- [Contributing](#contributing)\n- [License](#license)\n\n---\n\n## Quick Start\n\n```bash\n# Add this marketplace to Claude Code\n/plugin marketplace add team-attention/plugins-for-claude-natives\n\n# Install any plugin\n/plugin install <plugin-name>\n```\n\n---\n\n## Available Plugins\n\n| Plugin | Description |\n|--------|-------------|\n| [agent-council](./plugins/agent-council/) | Collect and synthesize opinions from multiple AI agents (Gemini, GPT, Codex) |\n| [clarify](./plugins/clarify/) | Transform vague requirements into precise specifications through iterative questioning |\n| [dev](./plugins/dev/) | Developer workflow: community opinion scanning and technical decision analysis |\n| [doubt](./plugins/doubt/) | Force Claude to re-validate its response when `!rv` is in your prompt |\n| [interactive-review](./plugins/interactive-review/) | Interactive markdown review with web UI for visual plan/document approval |\n| [say-summary](./plugins/say-summary/) | Speaks a short summary of Claude's response using macOS TTS (Korean/English) |\n| [youtube-digest](./plugins/youtube-digest/) | Summarize YouTube videos with transcript, insights, Korean translation, and quizzes |\n| [gmail](./plugins/gmail/) | Multi-account Gmail integration with email reading, searching, sending, and management |\n| [google-calendar](./plugins/google-calendar/) | Multi-account Google Calendar integration with parallel querying and conflict detection |\n| [kakaotalk](./plugins/kakaotalk/) | Send and read KakaoTalk messages on macOS using Accessibility API |\n| [session-wrap](./plugins/session-wrap/) | Session wrap-up, history analysis, and session validation toolkit |\n| [team-assemble](./plugins/team-assemble/) | Dynamically assemble expert agent teams for complex tasks using Claude Code's agent teams feature |\n| [podcast](./plugins/podcast/) | Generate Korean podcast episodes from any source with OpenAI TTS and YouTube auto-upload |\n| [fetch-tweet](./plugins/fetch-tweet/) | Fetch full tweet text, author info, and engagement data from X/Twitter URLs without authentication |\n\n## Plugin Details\n\n### agent-council\n\n![Demo](./assets/agent-council.gif)\n\n**Summon multiple AI models to debate your question and reach a consensus.**\n\nWhen you're facing a tough decision or want diverse perspectives, this plugin queries multiple AI agents (Gemini CLI, GPT, Codex) in parallel and synthesizes their opinions into a single, balanced answer.\n\n**Trigger phrases:**\n- \"summon the council\"\n- \"ask other AIs\"\n- \"what do other models think?\"\n\n**How it works:**\n1. Your question is sent to multiple AI agents simultaneously\n2. Each agent provides its perspective\n3. Claude synthesizes the responses into a consensus view with noted disagreements\n\n```bash\n# Example\nUser: \"summon the council - should I use TypeScript or JavaScript for my new project?\"\n```\n\n---\n\n### clarify\n\n![Demo](./assets/clarify.gif)\n\n**Turn vague requirements into precise, actionable specifications.**\n\nBefore writing code based on ambiguous instructions, this plugin conducts a structured interview to extract exactly what you need. No more assumptions, no more rework.\n\n**Trigger phrases:**\n- \"/clarify\"\n- \"clarify requirements\"\n- \"what do I mean by...\"\n\n**The process:**\n1. **Capture** - Record the original requirement verbatim\n2. **Question** - Ask targeted multiple-choice questions to resolve ambiguities\n3. **Compare** - Present before/after showing the transformation\n4. **Save** - Optionally save the clarified spec to a file\n\n**Example transformation:**\n\n| Before | After |\n|--------|-------|\n| \"Add a login feature\" | Goal: Add username/password login with self-registration. Scope: Login, logout, registration, password reset. Constraints: 24h session, bcrypt, rate limit 5 attempts. |\n\n---\n\n### dev\n\n**Developer workflow tools: community scanning and technical decision-making.**\n\nThis plugin provides two powerful skills for developer research and decision-making.\n\n#### Skills\n\n**`/dev-scan`** - Scan developer communities for real opinions\n- Searches Reddit (via Gemini CLI), Hacker News, Dev.to, and Lobsters in parallel\n- Synthesizes consensus, controversies, and notable perspectives\n- Great for understanding community sentiment before adopting a tool\n\n**`/tech-decision`** - Deep technical decision analysis\n- Multi-phase workflow with 4 specialized agents running in parallel\n- Combines codebase analysis, docs research, community opinions, and AI perspectives\n- Produces executive-summary-first reports with scored comparisons\n\n**Trigger phrases:**\n- \"developer reactions to...\", \"what do devs think about...\"\n- \"A vs B\", \"which library should I use\", \"기술 의사결정\"\n\n**How tech-decision works:**\n\n```\nPhase 1: Parallel Information Gathering\n┌─────────────────┬─────────────────┬─────────────────┬─────────────────┐\n│ codebase-       │ docs-           │ dev-scan        │ agent-council   │\n│ explorer        │ researcher      │ (community)     │ (AI experts)    │\n└────────┬────────┴────────┬────────┴────────┬────────┴────────┬────────┘\n         └─────────────────┴─────────────────┴─────────────────┘\n                                    │\nPhase 2: Analysis & Synthesis       ▼\n┌─────────────────────────────────────────────────────────────────────────┐\n│                        tradeoff-analyzer                                 │\n└─────────────────────────────────────────────────────────────────────────┘\n                                    │\n                                    ▼\n┌─────────────────────────────────────────────────────────────────────────┐\n│                       decision-synthesizer                               │\n│                    (Executive Summary First)                             │\n└─────────────────────────────────────────────────────────────────────────┘\n```\n\n```bash\n# Examples\nUser: \"React vs Vue for my new project?\"\nUser: \"Which state management library should I use?\"\nUser: \"Monolith vs microservices for our scale?\"\n```\n\n---\n\n### doubt\n\n**Force Claude to double-check its response before delivering.**\n\nWhen you include `!rv` anywhere in your prompt, Claude will pause before responding, re-validate its answer against potential errors, and only then deliver the response. Perfect for critical decisions or when you want extra confidence.\n\n**Trigger:**\n- Include `!rv` anywhere in your prompt\n\n**How it works:**\n1. `UserPromptSubmit` hook detects `!rv` keyword and sets a flag\n2. `Stop` hook intercepts Claude's response before delivery\n3. Claude re-validates the response for errors, hallucinations, or questionable claims\n4. Only after verification does Claude deliver the final answer\n\n**Why `!rv` instead of `!doubt`?**\nThe word \"doubt\" affects Claude's behavior - it starts doubting from the beginning. `!rv` (re-validate) is neutral.\n\n```bash\n# Example\nUser: \"What's the time complexity of binary search? !rv\"\n# Claude will verify its answer before responding\n```\n\n---\n\n### interactive-review\n\n**Review Claude's plans and documents through a visual web interface.**\n\nInstead of reading long markdown in the terminal, this plugin opens a browser-based UI where you can check/uncheck items, add comments, and submit structured feedback.\n\n**Trigger phrases:**\n- \"/review\"\n- \"review this plan\"\n- \"let me check this\"\n\n**The flow:**\n1. Claude generates a plan or document\n2. A web UI opens automatically in your browser\n3. Review each item with checkboxes and optional comments\n4. Click Submit to send structured feedback back to Claude\n5. Claude adjusts based on your approved/rejected items\n\n---\n\n### say-summary\n\n![Demo](./assets/say-summary.gif)\n\n**Hear Claude's responses spoken aloud (macOS only).**\n\nThis plugin uses a Stop hook to summarize Claude's response to a short headline and speaks it using macOS text-to-speech. Perfect for when you're coding and want audio feedback.\n\n**Features:**\n- Summarizes responses to 3-10 words using Claude Haiku\n- Auto-detects Korean vs English\n- Uses appropriate voice (Yuna for Korean, Samantha for English)\n- Runs in background, doesn't block Claude Code\n\n**Requirements:**\n- macOS (uses the `say` command)\n- Python 3.10+\n\n---\n\n### youtube-digest\n\n![Demo](./assets/youtube-digest.jpeg)\n\n**Summarize YouTube videos with transcripts, translations, and comprehension quizzes.**\n\nDrop a YouTube URL and get a complete breakdown: summary, key insights, full Korean translation of the transcript, and a 3-stage quiz (9 questions total) to test your understanding.\n\n**Trigger phrases:**\n- \"summarize this YouTube\"\n- \"digest this video\"\n- YouTube URL\n\n**What you get:**\n1. **Summary** - 3-5 sentence overview with key points\n2. **Insights** - Actionable takeaways and ideas\n3. **Full transcript** - With Korean translation and timestamps\n4. **3-stage quiz** - Basic, intermediate, and advanced questions\n5. **Deep Research** (optional) - Web search to expand on the topic\n\n**Output location:** `research/readings/youtube/YYYY-MM-DD-title.md`\n\n---\n\n### gmail\n\n<img src=\"https://upload.wikimedia.org/wikipedia/commons/thumb/7/7e/Gmail_icon_%282020%29.svg/1280px-Gmail_icon_%282020%29.svg.png\" width=\"120\" alt=\"Gmail\">\n\n**Manage multiple Gmail accounts from Claude Code.**\n\nRead, search, send, and manage emails across multiple Google accounts with full Gmail API integration.\n\n**Trigger phrases:**\n- \"check my email\"\n- \"send an email to...\"\n- \"search for emails from...\"\n- \"reply to this email\"\n- \"mark as read\"\n\n**Features:**\n- Multi-account support via `accounts.yaml` (work, personal, etc.)\n- Gmail search query syntax support\n- Email sending with attachments and HTML\n- Label and draft management\n- **5-step email sending workflow** with context gathering, draft review, and test delivery\n- Rate limiting and quota management\n- Batch processing and local caching\n\n**5-Step Email Workflow:**\n1. **Context gathering** - Parallel Explore agents search recipient info and related projects\n2. **Previous conversations** - Search recent emails to determine reply vs new thread\n3. **Draft composition** - Create draft with user feedback\n4. **Test send** - Send to your own email for verification\n5. **Actual send** - Deliver to recipient\n\n**Setup:**\n1. Create Google Cloud project with Gmail API enabled\n2. Run setup script for each account\n\n```bash\n# One-time setup per account\nuv run python scripts/setup_auth.py --account work\nuv run python scripts/setup_auth.py --account personal\n```\n\nAccount metadata is stored in `accounts.yaml` for easy management.\n\n---\n\n### google-calendar\n\n<img src=\"https://upload.wikimedia.org/wikipedia/commons/thumb/a/a5/Google_Calendar_icon_%282020%29.svg/960px-Google_Calendar_icon_%282020%29.svg.png\" width=\"120\" alt=\"Google Calendar\">\n\n**Manage multiple Google Calendar accounts from Claude Code.**\n\nQuery, create, update, and delete events across multiple Google accounts (work, personal, etc.) with automatic conflict detection.\n\n**Trigger phrases:**\n- \"show my schedule\"\n- \"what's on my calendar\"\n- \"create a meeting\"\n- \"check for conflicts\"\n\n**Features:**\n- Parallel querying across multiple accounts\n- Conflict detection between accounts\n- Full CRUD operations (create, read, update, delete)\n- Pre-authenticated with refresh tokens (no repeated logins)\n\n**Setup required:**\n1. Create Google Cloud project with Calendar API\n2. Run setup script for each account\n\n```bash\n# One-time setup per account\nuv run python scripts/setup_auth.py --account work\nuv run python scripts/setup_auth.py --account personal\n```\n\n---\n\n### kakaotalk\n\n![Demo](./assets/kakaotalk.gif)\n\n**Send and read KakaoTalk messages from Claude Code on macOS.**\n\nUses macOS Accessibility API to control the KakaoTalk app. Send messages or read chat history using natural language.\n\n**Trigger phrases:**\n- \"카톡 보내줘\", \"카카오톡 메시지\"\n- \"~에게 메시지 보내줘\"\n- \"채팅 읽어줘\"\n- \"KakaoTalk message\"\n\n**Features:**\n- Natural language message sending (with confirmation before send)\n- Chat history retrieval\n- Chat room listing\n- Auto-signature \"sent with claude code\"\n\n**Requirements:**\n- macOS only\n- KakaoTalk app must be running\n- Accessibility permission required\n\n```\n# Examples (natural language)\n\"구봉한테 밥 먹었어? 보내줘\"\n\"구봉이랑 대화 내역 보여줘\"\n```\n\n---\n\n### session-wrap\n\n**Comprehensive session wrap-up and analysis toolkit.**\n\nEnd your coding sessions with thorough analysis, and dive deep into session history for insights.\n\n#### Skills\n\n**`/wrap`** - Session wrap-up workflow\n- 2-phase multi-agent pipeline for comprehensive session analysis\n- Captures documentation needs, automation opportunities, learnings, and follow-ups\n- `/wrap [commit message]` for quick commits\n\n**`/history-insight`** - Session history analysis\n- Analyze Claude Code session history for patterns and insights\n- Search current project or all sessions\n- Extract themes, decisions, and recurring topics\n\n**`/session-analyzer`** - Post-hoc session validation\n- Validate session behavior against SKILL.md specifications\n- Check if agents, hooks, and tools executed correctly\n- Generate detailed compliance reports\n\n**How /wrap works (2-Phase Pipeline):**\n\n```\nPhase 1: Analysis (Parallel)\n┌──────────────┬──────────────┬──────────────┬──────────────┐\n│ doc-updater  │ automation-  │ learning-    │ followup-    │\n│              │ scout        │ extractor    │ suggester    │\n└──────┬───────┴──────┬───────┴──────┬───────┴──────┬───────┘\n       └──────────────┴──────────────┴──────────────┘\n                            │\nPhase 2: Validation         ▼\n┌─────────────────────────────────────────────────────────────┐\n│                    duplicate-checker                         │\n└─────────────────────────────────────────────────────────────┘\n                            │\n                            ▼\n                    User Selection\n```\n\n**Benefits:**\n- Never forget to document important discoveries\n- Identify patterns worth automating\n- Create clear handoff points for future sessions\n- Analyze past sessions for recurring patterns\n- Validate skill implementations against specifications\n\n---\n\n### team-assemble\n\n**Dynamically assemble expert agent teams for complex tasks.**\n\nInstead of manually designing agents, this plugin analyzes your task, scouts the codebase, and assembles an optimal team with the right roles, dependencies, and validation criteria — all using Claude Code's agent teams feature.\n\n> **Prerequisite:** Agent teams must be enabled. See [setup guide](./plugins/team-assemble/skills/team-assemble/references/enable-agent-teams.md).\n\n**Trigger phrases:**\n- \"assemble a team to...\"\n- \"team assemble\"\n- \"use a team for...\"\n\n**6-Phase Workflow:**\n\n```\nPhase 1 → Phase 2 → Phase 3 → Phase 4 → Phase 5 → Phase 6\nTask       Codebase   Integrate  Execute   Validate   Complete\nAnalysis   Scouts     & Confirm                       & Cleanup\n```\n\n**Key features:**\n- **Dynamic agent design** — scouts your codebase and proposes agents tailored to the task (no fixed catalog)\n- **Model 3-tier** — opus for strategy, sonnet for execution, haiku for research\n- **Parallel execution** — independent agents run simultaneously\n- **Acceptance criteria** — every team has measurable validation criteria\n- **Verify/fix loop** — QA validates, support fixes (max 3 rounds)\n- **Two approval gates** — confirm scope (Phase 1) and team composition (Phase 3)\n\n```bash\n# Examples\nUser: \"Assemble a team to refactor authentication from session-based to JWT\"\nUser: \"Use a team to evaluate Redis vs Memcached vs in-memory caching\"\nUser: \"Team assemble — extract shared utils from three microservices into a common lib\"\n```\n\n---\n\n### podcast\n\n**Turn any source into a Korean podcast episode, automatically uploaded to YouTube.**\n\nDrop URLs, tweets, articles, or PDFs and this plugin will analyze them, write a conversational Korean script, generate audio using OpenAI's `gpt-4o-mini-tts`, and upload the result to YouTube — all in one go.\n\n**Trigger phrases:**\n- \"make a podcast from this\"\n- \"팟캐스트 만들어\"\n- \"turn this into an episode\"\n- \"이 글을 팟캐스트로\"\n\n**The pipeline:**\n\n```\nSources → Parallel Analysis → Script Writing → TTS (OpenAI) → MP4 → YouTube Upload\n```\n\n**What you get:**\n1. **Script** - 8-12 min Korean podcast script with opening, analysis, fusion, and closing\n2. **Audio** - MP3 generated via OpenAI gpt-4o-mini-tts with natural Korean voice\n3. **Video** - MP4 with dark title card (1920x1080)\n4. **YouTube** - Auto-uploaded as unlisted with metadata\n\n**Partial execution supported:**\n- \"Just write the script\" → Script only\n- \"Generate TTS from this script\" → Audio only\n- \"Upload to YouTube\" → Upload existing MP4\n\n**Requirements:**\n- ffmpeg (for audio merging and MP4 conversion)\n- OpenAI API key (`OPENAI_API_KEY` env var)\n- Google OAuth client secret (for YouTube upload)\n\n```bash\n# Example\nUser: \"이 두 개의 아티클로 팟캐스트 만들어줘\"\n# → Analyzes both articles in parallel\n# → Writes fusion script\n# → Generates audio\n# → Uploads to YouTube\n```\n\n---\n\n### fetch-tweet\n\n**Read any public tweet from Claude Code — no auth, no API key, no JavaScript.**\n\nDrop an X/Twitter URL and get the full tweet text (URLs expanded), author info, engagement metrics, attached media, and quoted tweets. Powered by the open-source [FxEmbed](https://github.com/FxEmbed/FxEmbed) project.\n\n**Trigger phrases:**\n- \"트윗 가져와\", \"트윗 번역해줘\"\n- \"fetch this tweet\", \"translate this tweet\"\n- Any X/Twitter URL (`x.com`, `twitter.com`)\n\n**Why this exists:**\nX removed unauthenticated tweet embeds, so reading a tweet from a script normally requires an API key or browser automation. This plugin sidesteps that by routing through `api.fxtwitter.com` — the same backend that powers fxtwitter link previews on Discord/Telegram.\n\n**Features:**\n- Zero dependencies (Python stdlib only)\n- Full tweet data including quote tweets and media\n- `--json` mode for pipeline use with other skills\n- WebFetch fallback when script execution isn't available\n\n```bash\n# Direct script usage\npython scripts/fetch_tweet.py https://x.com/sama/status/...\npython scripts/fetch_tweet.py <url> --json | jq '.tweet.text'\n```\n\n**Limitations:** Private/deleted tweets cannot be fetched.\n\n---\n\n## Contributing\n\nContributions welcome! Please open an issue or PR.\n\n## License\n\nMIT\n",
  "bytes": 19262,
  "sha": "dc7d8f1bedbc20c690c2b7e0d23c13d64babe0ba85229ad7278655a1d5cb0a1f",
  "repo_slug": "team-attention/plugins-for-claude-natives",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/skl_team_attention_plugins_for_claude_native_79a4805b/readme"
}