The Discord voice exposes 11 tools (post, edit, delete, react, DM, list channels/members/messages). Write tools are marked destructive: true, so a HITL-enabled runtime gates every post/edit/delete behind operator approval. Useful for onboarding moderators without giving them direct delete rights.
Score file
import { defineScore, AnthropicProvider } from '@tuttiai/core'
import { DiscordVoice } from '@tuttiai/discord'
export default defineScore({
provider: new AnthropicProvider(),
agents: {
mod: {
name: 'mod',
model: 'claude-sonnet-4-6',
system_prompt:
'You are a community moderator. When users flag content, read the relevant messages, summarise, and propose moderation actions. Do NOT execute moderation actions without explicit operator approval.',
voices: [new DiscordVoice()],
permissions: ['network'],
},
},
}) How to run it
- 01 Scaffold a project:
npx tutti-ai init my-discord-mod - 02 Install voices:
npm i @tuttiai/discord - 03 Drop the score above into
tutti.score.ts, set the env vars in.env, and runtutti-ai run.
Why this is safe
-
HITL on every post/edit/delete/DM -
permissions: ["network"]