PR #80235: feat(discord): add implicit reply mention policy

• Repository: openclaw/openclaw
• Author: JiehoonKwak
• State: open | merged: False
• Link: https://github.com/openclaw/openclaw/pull/80235

Description (problem / solution / changelog)

Closes #80234.

Adds channels.discord.implicitReplyMentions:

{
  fromUsers: true,
  fromBots: true
}

Defaults preserve current behavior. Multi-bot deployments can set fromBots: false so bot-authored native replies do not wake another bot via the implicit reply_to_bot path; explicit <@botId> mentions still work.

Checks run:

• pnpm exec vitest run extensions/discord/src/monitor/message-handler.preflight.test.ts --maxWorkers=1
• pnpm config:schema:check
• pnpm config:docs:check
• pnpm check:bundled-channel-config-metadata
• git diff --check

Real behavior proof

Behavior or issue addressed: Discord native reply-only mentions can wake an OpenClaw bot when another bot replies to one of its messages. This PR adds a policy knob so deployments can keep human reply activation while disabling bot-authored reply-only activation.

Real environment tested: Private Discord server with a private test channel and two OpenClaw Discord bot accounts. Server, channel, bot names, message IDs, and account IDs are intentionally omitted from this public proof.

Exact steps or command run after this patch:

1. In the private test channel, send an anchor message from the target bot.
2. Send a native Discord reply from a separate relay bot to that anchor message. The reply body contains no textual <@bot> mention.
3. Observe whether the target bot wakes/responds.
4. Repeat after applying this branch with implicitReplyMentions: { fromUsers: true, fromBots: false }.
5. Send a separate control message from the relay bot that contains a real textual <@bot> mention and asks the target bot to reply with a known marker.

Evidence after fix: Redacted live output excerpt from the private Discord test channel:

after-fix native reply-only test:
relay bot | "oc-proof-80235-after4-231018 native-reply-no-explicit-mention" | ref=<target-bot-anchor>
UNEXPECTED_TARGET_RESPONSES=0

explicit mention control:
target bot | "oc-proof-80235-control-231107 explicit-mention-ok" | ref=<relay-bot-control-message>
EXPLICIT_CONTROL_OK=true

Observed result after fix: With fromBots:false, the bot-authored native reply-only message did not wake the target bot during the observation window. A real textual explicit mention still woke the target bot and produced the requested marker response.

What was not tested: Human-authored reply-only behavior was covered by unit tests but not repeated in the live Discord proof. PluralKit behavior was covered by unit tests but not repeated in the live Discord proof.

Changed files

• CHANGELOG.md (modified, +1/-0)
• docs/.generated/config-baseline.sha256 (modified, +3/-3)
• docs/channels/discord.md (modified, +23/-1)
• docs/channels/groups.md (modified, +1/-0)
• extensions/discord/src/monitor/message-handler.preflight-helpers.ts (modified, +11/-1)
• extensions/discord/src/monitor/message-handler.preflight.test-helpers.ts (modified, +2/-0)
• extensions/discord/src/monitor/message-handler.preflight.test.ts (modified, +189/-3)
• extensions/discord/src/monitor/message-handler.preflight.ts (modified, +24/-4)
• src/config/bundled-channel-config-metadata.generated.ts (modified, +16/-16)
• src/config/types.discord.ts (modified, +7/-0)
• src/config/zod-schema.providers-core.ts (modified, +8/-0)