The wizard has five steps: describe your community (text or voice, AI turns it into settings), choose a protection level (light, medium or strict), confirm the bot rights, set a link whitelist, and go live. Every choice becomes a normal setting you can fine-tune later in Settings — the wizard is a fast start, not a lock-in.
Want to set this up? Do it right in your Telm account.
Open in your dashboard1What the wizard does
The first time you open a group in the dashboard, the onboarding wizard tailors Telm to your community. It turns a few simple answers into a complete moderation configuration, so you never have to touch dozens of individual settings by hand to get started.
It writes a coherent set of group settings for you: it starts from a level preset, layers your explicit allow or block choices on top, and folds in any AI suggestions from your community description. The result is a single, consistent configuration rather than a pile of unrelated switches.
Everything it sets is a normal setting you can change afterwards — the wizard is a fast, guided start, not a one-way door.
2The five steps
The wizard guides you through five steps:
- About the group — describe your community in text or by voice; the AI turns it into sensible defaults.
- Protection level — pick how strict moderation should be: light, medium or strict.
- Bot rights — confirm the bot has Delete messages and Ban users.
- Link whitelist — list domains that are always allowed (your site, docs, socials).
- Go live — review and switch protection on.
3Step 1 - describe your community
The first step asks you to tell Telm what your group is about, in a sentence or two. You can type it, or tap the microphone and record a short voice note — the note is transcribed and analysed the same way as typed text.
The AI reads your description and pre-tunes the defaults so they fit your community. For a crypto trading chat it might expect and allow finance talk while still blocking P2P scam offers; for a marketplace it can permit genuine sales posts that a generic filter would flag; for a support group it can be stricter about outside promotion. This is why two groups on the same level can end up with slightly different, better-fitting settings.
This step is optional. Skip it and you simply get the clean level preset with no AI adjustments. You can also choose whether to save your description as reusable AI context for other features.
4Step 2 - choose a protection level
The level is the backbone of your configuration. All three levels always run the core engine — the CAS blocklist and AI moderation are on at every level, . The levels differ in how many mechanical restrictions they add on top.
- Light — AI and CAS only, no mechanical restrictions. Members chat freely with the fewest false positives. Best for small, trusted or invite-only communities.
- Medium — closes the common spam vectors: outside promotion, DM-invite lures, P2P crypto offers, duplicate flooding and raids, with an anti-raid shield that challenges suspicious joins. A balanced default for most public groups.
- Strict — a fortress: captcha for new members, a 24-hour newcomer quarantine, links allowed only from your whitelist, rate limiting against flooding, and automatic lockdown during a raid. Best for large public groups under heavy spam pressure.
5Steps 3 and 4 - rights and whitelist
Step three confirms the bot has the two rights it needs — Delete messages and Ban users. If either is missing, the wizard tells you exactly what to enable so the level you chose can actually be enforced. This is the same requirement covered in Setting up admin rights.
Step four sets your link whitelist: the domains that are always allowed, no matter how strict your link policy is. Add your own website, documentation, blog and official social profiles here so your own links are never caught. On the strict level, where links are otherwise limited to the whitelist, this step matters most — it is what keeps your own content flowing while blocking everything else.
You can also fine-tune what is allowed versus blocked on these steps — for example permitting media and stickers while blocking forwards and DM invites — before you commit.
6Step 5 - go live (or watch first)
The final step shows a summary of everything the wizard is about to apply and asks you to confirm. When you finish, the settings are saved and, if you choose to activate protection, the bot begins enforcing them immediately.
If you would rather see the bot in action before it acts on anyone, leave protection in observation and turn on monitoring mode. The bot then logs every decision it would have made — what it would delete, whom it would restrict — without touching a single message. Review those logs for a day, and when you are comfortable, switch enforcement on with one click.
7After the wizard
Once you finish, the group is protected and every choice becomes an ordinary setting. From here you can open any section of the dashboard to fine-tune filters, new-member handling, welcome messages, custom rules and more.
Common next steps are setting up a welcome message, adding your own custom rules for community-specific phrases, and inviting your moderator team. You can also re-run the wizard later if your community changes character.
Nothing here is permanent: adjust a single toggle, change your level, or re-run onboarding whenever you like.