An autonomous QA agent that tests your web app in a real browser, remembers how your product behaves, and tells you the moment something silently breaks.
Replies
Best
Maker
📌
I'm Anoop, maker of Argus.
This started with a pattern I kept seeing everywhere: every team has one QA person who knows everything — how
checkout behaves for guest users, why that button shows no toast, what broke last release. When they're on
leave, the whole team is guessing. And the worst bugs never crash anything — the page renders fine while the API
behind it quietly returns 500s. Nobody notices until a customer does.
So I built Argus — the hundred-eyed watchman from Greek myth, reborn as a QA agent:
Point it at your web app — it explores it in a real browser and writes a living spec of how your product
actually behaves
Every night it re-walks your flows and flags pass / drift / broken — "drift" is the "works, but different
than before" category most tools miss
It watches every API call and console error while clicking — silent backend failures become findings with the
endpoint, status, and response body quoted
Findings come with screenshot + repro steps, and file to Jira in one click
And it remembers everything — ask "how does X work?" months later and it answers from verified memory, not
tribal knowledge
A few things that matter to me:
— Runs 100% on your machine. Your data never touches our servers (there are none).
— Bring your own AI key — Claude, GPT, or Gemini.
— The agent never sees your passwords — logins are stored encrypted and replayed by name; SSO sessions you
capture once in a real browser window.
— It asks before doing anything destructive. Always.
It's early days and that's the honest pitch: I want you to break it. If Argus misses a bug in your app — or you
find one in Argus itself — I want to hear about it. A bug-finder deserves to be bug-hunted. 🙂
Question for you all: what's the one flow in your product you're most scared of silently breaking?
Replies