@slack/web-api@7.14.0

Agent Thinking Steps: Display Tasks/Tools, Plans, and Markdown Text

🍿 Preview: Display as Plan

https://github.com/user-attachments/assets/1a6478c2-37aa-48c2-84e9-69375b20177e

🍿 Preview: Display as Timeline

https://github.com/user-attachments/assets/984f84e5-c0e0-47ef-a1c2-994ac51a25ba

📺 Chat Stream with 2 Display Mode

• Plan Display Mode
• Timeline Display Mode

👾 Chat Stream Structured Content

Now, you can display a mixture of structured content called "chunks":

• 🔠 Markdown Text Block to format your text with standard markdown
• ☑️ Task Card Block to display a single task, representing an AI Tool Call or general action
• 🗒️ Plan Block to display a collection of related tasks
• 📚 URL Sources Element to display references within a task card block

Available in:

• 🔌 API Methods: chat.startStream, chat.appendStream, and chat.stopStream
• 🛟 Chat Stream Helper: const stream = new ChatStreamer(...);, stream.append(...)

📖 Documentation

• Announcements
  • CHANGELOG
• Guides
  • AI in Slack Apps
  • Developing AI Apps (streaming section)
  • Sending and Scheduling Messages (streaming section)
• Block Kit
  • Plan Block
  • Task Card Block
  • URL Source Element
• References
  • API Methods: chat.startStream, chat.appendStream, and chat.stopStream
  • Node SDK Reference

🍿 Getting Started

$ slack create

# → AI Agent App
#   → Bolt for JavaScript
#     Bolt for Python

---

Minor Changes

• 1fbce32: feat: add thinking steps support to streaming methods

  chat.appendStream, chat.startStream, and chat.stopStream now accept a chunks parameter for streaming structured content including markdown text, plan updates, and task updates.

  Related PRs:

  • #2467 - accept chunks as arguments to chat.{start,append,stop}Stream methods
  • #2470 - accept chunks as arguments to chat stream helper
  • #2479 - add task display mode option to start of chat streams
  • #2481 - export the chat streamer and related options from the package

  Example

  const stream = new ChatStreamer(client, client.logger, {
    channel: CHANNEL_ID,
    thread_ts: threadTs,
  });
  
  await stream.append({
    chunks: [
      {
        type: "markdown_text",
        text: "**Hello!** I am starting to process your request...\n\n",
      },
    ],
  });
  
  await stream.append({
    chunks: [
      {
        type: "plan_update",
        title: "Processing tasks...",
      },
      {
        type: "task_update",
        id: "task-1",
        title: "Fetching data from API",
        status: "complete",
        output: "Successfully retrieved 42 records",
      },
    ],
  });
  
  await stream.stop({
    chunks: [
      {
        type: "markdown_text",
        text: "\n\n---\n\n✅ **All tasks completed successfully!**\n",
      },
    ],
  });

Patch Changes

• 16a43ca: fix(web-api): add channel_id to canvases.create method
• Updated dependencies [f1fb7bf]
  • @slack/types@2.20.0