Is “teammate” the same thing as “AI agent”?

Yes. “AI agent” is the technical term and “teammate” is what it feels like to work with one. We use both, and they mean the same thing.

Can I have more than one?

[PLACEHOLDER: confirm whether one person can run several projects, each with its own agent.]

Does it work with my human team?

Yes. It works alongside people. Think of it as the teammate who handles the repeating work so your people can do the rest.

What does it do without asking me?

[CONFIRM: what the agent does on its own versus what needs approval.]

Can I change its personality later?

Yes. You can change how your project sounds at any point, and your teammate carries the change everywhere it shows up.

What if it says something daft to my community?

You can delete the message, correct what it knows so it does not repeat the mistake, and turn approvals back on if you want to read everything first. People can see it is a bot, so a mistake reads as a bot mistake, not as you. [CONFIRM AGAINST PRODUCT]

If I stop using launchd, do I lose my group?

No. Your Telegram group or channel is yours and stays yours — it was never ours. What you would lose is the teammate working in it. [CONFIRM]

What is a project agent?

A project agent is an AI agent dedicated to a single project. It has one goal, growing that project, and it works toward it every day alongside the people behind it. On launchd, every project gets one. It lives in Telegram as the project’s own bot, with its own name and personality. We call it your first teammate.

Launch your thing What it does

Personality included.

How it’s different

Not like

A general AI assistant

Helps you with anything. Starts fresh each time. Works for everyone.

A project agent

Works on one project. Remembers all of it. Works only for you.

Not like

A normal Telegram bot

Responds to commands. Does exactly what it was programmed to do.

A project agent

Decides and suggests. Acts when nobody has said anything.

What it does, concretely

[CONFIRM AGAINST PRODUCT]

Writes and schedules the announcements and updates you keep meaning to write
Welcomes newcomers and answers their questions in your project’s voice
Keeps your group or channel active when you are busy or asleep
Suggests what to do next, every day, based on what is working
Tracks your launch plan and tells you when something has slipped
Hands things to your human team when they need a person

Brain and personality

Two things make your teammate yours: what it knows, and how it sounds.

The brain is everything you tell it about the project — what you are making, who it is for, what you would never say. The personality is the voice you choose: funny, serious, calm, loud. It stays in character everywhere it shows up, and you can change both whenever you like.

Who’s in control

You are. Your teammate proposes, you decide what it can do on its own. Start with everything needing your approval, then hand over the jobs you stop wanting to read.

[CONFIRM: exact split between autonomous and approved actions.]

What it can’t do

It cannot make your project good. If the thing you built is not useful, no amount of posting fixes that.
It cannot meet anyone in person, get on a call, or shake a hand.
It cannot be accountable. When something goes wrong, it is your name on it.
It cannot know what it was not told. A vague brief produces vague work.
It can be wrong, confidently. That is why you can review what it does.

When it gets something wrong

It will, eventually. Saying “it can be wrong” and stopping there is not good enough, so here is what actually happens next.

People know it is a bot.

Your teammate is your project’s bot and Telegram labels it as one. It never pretends to be you or to be a person. So a mistake reads as a bot getting something wrong, not as you saying something strange.

You can delete what it said.

It posts as your project’s bot in your group, and you can remove any message it sent, the same way you would remove your own. [CONFIRM AGAINST PRODUCT]

You correct it once, not every time.

Telling it what it got wrong updates what it knows about your project, so the same mistake does not come back next week. [CONFIRM AGAINST PRODUCT]

You can put it on a short leash.

If it has lost your trust, turn its approvals back on and read everything before it goes out. You can loosen that again whenever you like. [CONFIRM AGAINST PRODUCT]

It should say so.

When your teammate is corrected in front of people, it says it got it wrong, plainly, in your project’s voice. No quiet edits. Your community can see what happened.

Questions

What is a project agent?: A project agent is an AI agent dedicated to a single project, with one goal: growing that project. On launchd it lives in Telegram as your project’s own bot.
Is “teammate” the same thing as “AI agent”?: Yes. “AI agent” is the technical term and “teammate” is what it feels like to work with one. We use both, and they mean the same thing.
Can I have more than one?: [PLACEHOLDER: confirm whether one person can run several projects, each with its own agent.]
Does it work with my human team?: Yes. It works alongside people. Think of it as the teammate who handles the repeating work so your people can do the rest.
What does it do without asking me?: [CONFIRM: what the agent does on its own versus what needs approval.]
Can I change its personality later?: Yes. You can change how your project sounds at any point, and your teammate carries the change everywhere it shows up.
What if it says something daft to my community?: You can delete the message, correct what it knows so it does not repeat the mistake, and turn approvals back on if you want to read everything first. People can see it is a bot, so a mistake reads as a bot mistake, not as you. [CONFIRM AGAINST PRODUCT]
If I stop using launchd, do I lose my group?: No. Your Telegram group or channel is yours and stays yours — it was never ours. What you would lose is the teammate working in it. [CONFIRM]