MMaia Docs
Product overview
Use and extend

Connect messaging platforms

Plan channel credentials, private-message behavior, identity mapping, and rollout tests.

Messaging platforms give users a familiar interface while Maia remains a single governed runtime. Each platform authenticates its sender differently, but Maia converts the result into a stable platform:user_id before authorization.

Choose a channel pattern

PatternUse it whenMain concern
Private or direct messageThe conversation can contain user-specific or sensitive context.Verify the sender's stable platform identity.
Restricted team channelSeveral governed users share an operational context.Do not treat channel membership as a Maia role.
Mention-based public channelMaia should respond only when explicitly invoked.Minimize exposed tools and prevent sensitive output.
Webhook or APIAnother system starts deterministic workflows.Authenticate requests and define a non-human governance identity.

Common platform families

Maia includes adapters for platforms such as Discord, Slack, Telegram, WhatsApp, Signal, Matrix, Mattermost, email, SMS, Home Assistant, Microsoft Teams, and webhook/API integrations. The installed version is authoritative because adapters and provider requirements evolve.

Prepare credentials

  1. Create a dedicated bot or application for the Maia deployment.
  2. Grant only the platform permissions required for the intended channels and message types.
  3. Store tokens through Maia's secret-aware Gateway or Keys interface.
  4. Record the bot owner, renewal process, permitted workspaces or servers, and incident-revocation procedure.
Do not reuse an upstream Hermes bot.

If Maia and Hermes use the same bot credential, both gateways can consume or reply to the same events. Give each runtime its own bot identity.

Roll out a platform safely

  1. Connect the bot to a private test channel.
  2. Admit the bootstrap administrator and verify a simple response.
  3. Add a second identity and confirm that it remains Pending Governance.
  4. Assign that identity a minimal role and verify /whoami.
  5. Test normal chat, an allowed tool, a denied resource, and any expected approval flow.
  6. Expand to production channels only after logs and audit events match the intended policy.

Apply channel safety

  • Require direct messages for dashboard sign-in tokens and other private responses.
  • Do not expose administration commands in broad channels.
  • Restrict file and terminal tools independently of channel membership.
  • Define how message history, attachments, and generated files are retained by the platform.
  • Use platform-native audit and revocation features alongside Maia's audit trail.
navigateEsc close