WeChat iLink Typing Indicator API — undocumented but works

[btxro3han-code]

TL;DR

You can show the native "对方正在输入..." (typing indicator) in WeChat using the iLink bot API. This is undocumented but fully functional. We implemented it in our WeChat bot and it works great — much better than sending a temporary "thinking..." message.

The API

Step 1: Get a typing_ticket

POST {baseUrl}/ilink/bot/getconfig

Headers:
  Content-Type: application/json
  AuthorizationType: ilink_bot_token
  Authorization: Bearer {bot_token}
  X-WECHAT-UIN: {base64(randomUint32())}

Body:
{
  "ilink_user_id": "{user_id}",     // ⚠️ REQUIRED — omitting this returns ret:-2
  "base_info": { "channel_version": "2.0.1" }
}

Response:
{
  "ret": 0,
  "typing_ticket": "AARTBK8FAAAB..."
}

The ticket can be cached for ~20 hours.

Step 2: Send/cancel typing status

POST {baseUrl}/ilink/bot/sendtyping

Body:
{
  "ilink_user_id": "{user_id}",     // ⚠️ REQUIRED here too
  "to_user_id": "{user_id}",
  "typing_ticket": "{ticket}",
  "command": 1,                      // 1 = show typing, 2 = cancel
  "base_info": { "channel_version": "2.0.1" }
}

Response:
{ "ret": 0 }

Key Gotcha

Both getconfig and sendtyping require the ilink_user_id field. Without it you get {"ret":-2,"errmsg":"ilink_user_id required"}. This is not documented anywhere we could find — we discovered it by trial and error.

TypeScript Implementation

// Typing ticket cache
let typingTicket: string | null = null;
let typingTicketExpiry = 0;

async function ensureTypingTicket(
  baseUrl: string, token: string, userId: string
): Promise<string | null> {
  if (typingTicket && Date.now() < typingTicketExpiry) return typingTicket;

  const res = await fetch(`${baseUrl}/ilink/bot/getconfig`, {
    method: 'POST',
    headers: buildIlinkHeaders(token),
    body: JSON.stringify({
      ilink_user_id: userId,
      base_info: { channel_version: '2.0.1' },
    }),
  });

  const data = await res.json();
  if (data.typing_ticket) {
    typingTicket = data.typing_ticket;
    typingTicketExpiry = Date.now() + 20 * 3600_000;
    return typingTicket;
  }
  return null;
}

async function sendTyping(
  baseUrl: string, token: string, userId: string, command: 1 | 2 = 1
) {
  const ticket = await ensureTypingTicket(baseUrl, token, userId);
  if (!ticket) return;

  await fetch(`${baseUrl}/ilink/bot/sendtyping`, {
    method: 'POST',
    headers: buildIlinkHeaders(token),
    body: JSON.stringify({
      ilink_user_id: userId,
      to_user_id: userId,
      typing_ticket: ticket,
      command,
      base_info: { channel_version: '2.0.1' },
    }),
  });
}

Usage Pattern

// When a message arrives:
await sendTyping(baseUrl, token, userId, 1);  // Show typing

// Process with AI...
const response = await queryAI(message);

await sendTyping(baseUrl, token, userId, 2);  // Cancel typing
await sendMessage(baseUrl, token, userId, response);

Why This Matters

Most WeChat bots either:

• Show no feedback while processing (user wonders if bot is alive)
• Send a temporary "thinking..." message (creates noise in chat history)

The native typing indicator solves both — user sees the bot is working, no extra messages.

Hope this helps anyone building WeChat bots with iLink! 🎉