WhatsApp group messages

For the cross-channel groups model (Discord, iMessage, Matrix, Microsoft Teams, QQBot, Signal, Slack, Telegram, WhatsApp, Zalo), see Groups. This page covers the WhatsApp-specific behavior on top of that model: activation, group allowlists, per-group session keys, and pending-message context injection.

Goal: let OpenClaw sit in WhatsApp groups, wake up only when pinged, and keep that thread separate from the personal DM session.

agents.entries.*.groupChat.mentionPatterns is shared with the other channels’ mention gating. For multi-agent setups, set it per agent, or use messages.groupChat.mentionPatterns as a global fallback. With neither set, patterns are derived from the agent identity name/emoji.

Behavior

• Activation modes: mention (default) or always. mention requires a ping: a real WhatsApp @-mention (mentionedJids), a configured regex pattern, the bot’s E.164 digits anywhere in the text, or a quoted reply to one of the bot’s messages (except shared-number self-chat setups). always wakes the agent on every admitted message and requires a reply by default. To permit selective silence for unaddressed messages, explicitly set surfaces.whatsapp.silentReply.group: "allow"; the automatic-reply prompt then permits the exact silent token NO_REPLY (case-insensitive). See Silent replies for the global setting and precedence. Activation defaults come from config (channels.whatsapp.groups requireMention) and can be overridden per group via /activation.
• Group allowlist: when channels.whatsapp.groups is set, only listed group JIDs are admitted (include "*" to allow all); messages from unlisted groups are dropped with a log hint.
• Group policy: channels.whatsapp.groupPolicy controls whether group messages are accepted (open|disabled|allowlist). allowlist uses channels.whatsapp.groupAllowFrom (fallback: explicit channels.whatsapp.allowFrom). Default is allowlist (blocked until you add senders).
• Per-group sessions: session keys look like agent:<agentId>:whatsapp:group:<jid> (non-default accounts append :thread:whatsapp-account-<accountId>), so directives such as /verbose on, /trace on, or /think high (sent as standalone messages) are scoped to that group; personal DM state is untouched.
• Context injection: pending-only group messages (default 50) that did not trigger a run are prefixed under [Chat messages since your last reply - for context], with the triggering line under [Current message - respond to this]. The pending window is cleared after the run; messages already in the session are not re-injected.
• Sender attribution: each group line carries the sender label inside the message envelope, e.g. [WhatsApp <groupJid> <timestamp>] Alice ([phone]): text, and sender identity plus group subject/members ride along in the untrusted conversation-metadata block.
• Ephemeral/view-once: wrappers are unwrapped before extracting text/mentions, so pings inside them still trigger.
• Group system prompt: the first turn of a group session (and any turn after /activation changes the mode) injects activation guidance into the system prompt (Activation: trigger-only ... or Activation: always-on ..., plus “address the specific sender”). Persistent group-chat delivery guidance (“You are in a WhatsApp group chat…”) is always included.

Config example (WhatsApp)

Make display-name pings work even when WhatsApp strips the visual @ from the text body:

{
  channels: {
    whatsapp: {
      groups: {
        "*": { requireMention: true },
      },
      historyLimit: 50, // pending group context window (default 50)
    },
  },
  agents: {
    entries: {
      main: {
        default: true,
        groupChat: {
          mentionPatterns: ["@?openclaw", "\\+?15555550123"],
        },
      },
    },
  },
}

Notes:

• The regexes are case-insensitive and use the same safe-regex guardrails as other config regex surfaces; invalid patterns and unsafe nested repetition are ignored.
• WhatsApp still sends canonical mentions via mentionedJids when someone taps the contact, so the number fallback is rarely needed but is a useful safety net.
• The pending-context window resolves as channels.whatsapp.accounts.<id>.historyLimit → channels.whatsapp.historyLimit → messages.groupChat.historyLimit → 50.

Activation command (owner-only)

Use the group chat command:

• /activation mention
• /activation always

Only owner numbers (from channels.whatsapp.allowFrom, or the bot’s own E.164 when unset) can change this; /activation from anyone else is ignored and stored as context only. Send /status as a standalone message in the group to see the current activation mode.

How to use

1. Add your WhatsApp account (the one running OpenClaw) to the group.
2. Say @openclaw ... (or include the number). Only allowlisted senders can trigger it unless you set groupPolicy: "open".
3. The agent prompt includes the pending group context plus sender-labeled lines so it can address the right person.
4. Session directives (/verbose on, /trace on, /think high, /new or /reset, /compact) apply only to that group’s session; send them as standalone messages so they register. Your personal DM session stays independent.

Testing / verification

• Manual smoke:
  • Send an @openclaw ping in the group and confirm a reply that references the sender name.
  • Send a second ping and verify the history block is included, then cleared on the next turn.
• Check gateway logs (run with --verbose) for inbound web message entries showing from: <groupJid> and the sender-labeled body.

Known considerations

• Heartbeats run in the agent’s main session; group sessions never get heartbeat runs.
• Echo suppression remembers the combined prompt (history + current message) per session so the bot’s own delivered messages do not retrigger it; an identical repeated batch can be skipped as an echo.
• Session store entries appear as agent:<agentId>:whatsapp:group:<jid> in the per-agent SQLite session store; a missing entry just means the group has not triggered a run yet.
• Typing indicators follow agents.entries.*.typingMode / agents.defaults.typingMode. When visible replies are opted into message-tool-only mode, typing starts immediately by default so group members can see the agent working even if no automatic final reply is posted. Explicit typing-mode config still wins.

Related

• Groups
• Channel routing
• Broadcast groups