Overview
Transition Timing controls when the AI begins steering toward paid content. Auto reads the full conversation and decides naturally. Fixed sets a latest boundary from 3 to 50 visible AI reply batches, while still allowing an earlier transition when the conversation clearly supports it.
- Let the AI pace neutral, flirty, and highly engaged fans differently.
- Set a latest commercial-transition boundary when a predictable deadline is needed.
- Understand that one multi-bubble AI reply still counts as one reply batch.
Before you start
- Select the exact Telegram bot and creator profile you want to configure.
- Make sure the bot has a usable Secretary Mode setup for live creator-profile conversations.
- Prepare at least one active, usable Script before enabling AI for production traffic.
Step by step
- Open Behavior Settings for the selected bot.
- Choose Auto when the AI should read the relationship and decide the moment itself.
- Choose Fixed when you want a latest boundary, then enter a value from 3 to 50.
- Test a neutral conversation and an engaged conversation with the same bot.
- Confirm the AI can begin earlier in Fixed when the exact commercial beat is already coherent.
How it works
These are the product rules and states that matter for this workflow.
| What to know | Telestars behavior |
|---|---|
| Auto | The AI continuously reassesses the newest fan meaning, relationship, memory, Voice Preset, Creator Information, and available Script context. |
| Fixed | The saved number is the latest point for beginning the transition, not a minimum wait and not a required media send on that exact reply. |
| Count unit | One successfully visible automated AI response batch counts once, even when it contains two chat bubbles. |
| Product limits | Transition Timing never bypasses Script order, media readiness, Waiting Payment, Waiting Interval, takeover, purchase, or delivery truth. |
Verify the result
- Send a real fan message to the linked Telegram profile and confirm a visible reply arrives.
- Confirm the reply uses the intended creator facts, voice, language, and commercial pace.
- Check that changing style never changes the configured Script, price, media, or delivery state.
Troubleshooting
| If you see this | What to check |
|---|---|
| Auto feels too slow | Check that a usable Script and factual media context exist, then review the whole conversation rather than a single fan phrase. |
| Fixed did not send media at the number | Fixed starts steering by the boundary; exact delivery still needs a coherent legal Script event and working Telegram media. |
| The AI counted too quickly | Confirm you are counting visible AI response batches, not individual bubbles or fan messages. |
Best practice
Auto is the LLM-first default. Fixed is useful when you want a deadline without scripting the conversation turn by turn.