Question 1

How does aImsg access my code?

Accepted Answer

Through GitHub OAuth — you authorize exactly which repos we can see. We read code directly from GitHub, process it in memory, and never store it. Your API keys are encrypted with AES-256-GCM. We're paranoid about security so you don't have to be.

Question 2

Which AI model does it use?

Accepted Answer

Whichever one you want. aImsg supports Claude, GPT, and Gemini. You bring your own API key (BYOK), so you control the model, the cost, and the context window. Swap providers anytime.

Question 3

Does it work with private repos?

Accepted Answer

Absolutely. Public, private, whatever — if you can see it on GitHub, aImsg can work with it. You pick which repos to authorize during OAuth setup. We never touch anything you haven't explicitly allowed.

Question 4

How do I get started?

Accepted Answer

Head to /start, verify your phone number, link GitHub, pick a repo, and drop in an API key. Takes about 2 minutes. Then just text us a question about your code.

Question 5

What can I actually do?

Accepted Answer

Eight commands: /help, /link, /repo, /status, /review [PR#], /ask <question>, /edit [path], /pr "title". Or just type in plain English — "fix the typo in README" works too.

Question 6

Is there a message limit?

Accepted Answer

You get 10 free messages to try it. After that, Starter ($5/mo) gives you 500 messages/month with a 50/day cap. Pro ($20/mo) is unlimited messages with a 200/day cap.

Question 7

Is my code safe?

Accepted Answer

Very. GitHub OAuth (we never see your password), AES-256-GCM for API keys, webhook signature verification, rate limiting, and read-only access by default. Every code change goes to a new branch — your main branch is never touched directly. You approve everything before it ships.

Question 8

Can I self-host it?

Accepted Answer

Not yet. aImsg is hosted — but it means zero infrastructure for you to manage.

Simple, transparent pricing

Starter

Pro

Things people actually ask