Go to console.cloud.google.com, create a new project (or select existing).

• Enable Google Chat API.
• Go to “APIs & Services” → “Credentials” → Create OAuth 2.0 Client ID (application type: “Desktop app” or “Web application” — for Apps Script we’ll use script’s built-in auth later).
• Enable Gemini API (or Vertex AI) if using Google AI. Otherwise grab Claude key separately.

Open script.google.com, start a new project. Name it “Chat summarizer bot”.

🔧 code overview (copy & customize)

// Google Chat AI summarizer — fetches thread, calls Gemini, posts summary

const GEMINI_API_KEY = "YOUR_GOOGLE_AI_STUDIO_KEY";  // from aistudio.google.com

function doGet(e) {
  return ContentService.createTextOutput("Bot is alive.");
}

// this function will be called when the bot is mentioned or slash command
function onMessage(event) {
  // event.space.name, event.message.text, event.message.thread
  if (event.message.slashCommand) {
    if (event.message.slashCommand.commandId === 1) { // /summarize
      return summarizeThread(event);
    }
  }
  return { text: "👋 use /summarize to get a summary of this thread." };
}

function summarizeThread(event) {
  const spaceName = event.space.name;
  const threadKey = event.message.thread.name; // unique thread id

  // 1. fetch recent messages in the thread (last 50)
  const messages = listMessages(spaceName, threadKey, 50);
  if (!messages || messages.length === 0) return { text: "no messages found." };

  // 2. format them into a prompt
  const threadText = messages.map(m => m.text).join("\n");
  const prompt = `Summarize the following Google Chat thread. Keep it friendly but concise.
  - highlight key decisions ✅
  - action items with owners (if mentioned) 📋
  - any deadlines or dates 📅
  - overall vibe / conclusion.

Thread messages:\n${threadText}\n\nSummary:`;

  // 3. call Gemini (or Claude)
  const summary = callGemini(prompt);

  // 4. post summary back to thread
  postMessage(spaceName, threadKey, `📋 *Thread summary:*\n${summary}`);
  return { text: "summary posted above 👆" };  // fallback
}

// helper: list messages from a thread
function listMessages(spaceName, threadKey, max=50) {
  const url = `https://chat.googleapis.com/v1/${spaceName}/messages?pageSize=${max}&threadKey=${encodeURIComponent(threadKey)}`;
  const response = UrlFetchApp.fetch(url, {
    headers: { Authorization: 'Bearer ' + ScriptApp.getOAuthToken() }
  });
  const data = JSON.parse(response);
  return data.messages || [];
}

// helper: call Gemini 1.5 Flash
function callGemini(promptText) {
  const url = `https://generativelanguage.googleapis.com/v1beta/models/gemini-1.5-flash:generateContent?key=${GEMINI_API_KEY}`;
  const payload = {
    contents: [{ parts: [{ text: promptText }] }],
    generationConfig: { temperature: 0.3, maxOutputTokens: 400 }
  };
  const options = {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    payload: JSON.stringify(payload)
  };
  const response = UrlFetchApp.fetch(url, options);
  const json = JSON.parse(response);
  return json.candidates[0].content.parts[0].text;
}

function postMessage(spaceName, threadKey, text) {
  const url = `https://chat.googleapis.com/v1/${spaceName}/messages`;
  const payload = {
    text: text,
    thread: { name: threadKey }
  };
  UrlFetchApp.fetch(url, {
    method: 'POST',
    headers: { Authorization: 'Bearer ' + ScriptApp.getOAuthToken() },
    contentType: 'application/json',
    payload: JSON.stringify(payload)
  });
}

After pasting, replace GEMINI_API_KEY with your actual key (from aistudio.google.com → Get API key).

• In Apps Script editor, click Project Settings → check “Show ‘appsscript.json’ manifest file in editor”.
• Go back to editor, open appsscript.json and add Chat scopes:

{
  "timeZone": "America/New_York",
  "dependencies": {},
  "exceptionLogging": "STACKDRIVER",
  "oauthScopes": [
    "https://www.googleapis.com/auth/chat.messages",
    "https://www.googleapis.com/auth/chat.spaces.readonly",
    "https://www.googleapis.com/auth/script.external_request"
  ],
  "chat": {
    "name": "Thread Summarizer",
    "description": "Summarizes long threads using AI",
    "slashCommands": [{
      "commandId": 1,
      "name": "summarize",
      "description": "Get a summary of this thread",
      "commandType": "MESSAGE"
    }]
  }
}

Save. Then deploy: Deploy → New deployment → type “Google Chat app”. Copy the deployment ID.

Go back to Google Cloud Console → APIs & Services → Google Chat API → Configuration.

• App name: Thread Summarizer
• Avatar: optional
• Description: friendly AI that summarizes threads
• Functionality: Check “Bot works in direct messages, room mentions, and space mentions”.
• Connection settings: “Apps Script project” → paste your Deployment ID.
• Slash commands: add command “/summarize” with ID 1.
• Permissions: “Specific people and groups in your domain” (or make it public internally).

Save. Now add the bot to any Chat space by @mentioning it using its email (looks like @threadsummarizer).

Go to a Google Chat space with a long thread. Type /summarize (or @bot /summarize). Wait a few seconds — the bot will fetch recent messages and post a concise summary right in the thread.

If something fails, check Apps Script Executions tab for logs.