Viktor Fasi

@monta Thanks a lot Monta, really appreciate the feedback.

@luki123code Yes, thanks for checking! For animated ui you can decide if you are freezing it taking screenshots some time apart or recording a video. For static pagic it's a lot simpler :)

I never trusted AI, even if I let it code and ship without reviewing the code. Why would I trust AI written tests? Manual testing becomes more and more of a strain, so I came up with Lastest.

"Boring CHANGELOG.md → stunning website" is exactly the right level of scope. Most changelog tools want to own the whole release-notes workflow; treating Markdown as the source of truth means it fits whatever process a team already has. One thing I'd want to know: does it watch the file directly (via Git/webhook on tag), or is publishing manual? Big difference for "fast-shipping teams."

"Replace the PDF pentest report" is the right enemy to pick. A 60-page PDF that nobody reads after the first week is a real problem. The question I'd want answered up front: how do findings flow into our existing tracker (Linear, Jira, GitHub Issues) so they become work, not artifacts? If that loop is closed, the human-powered angle is a much easier sell than another scanner.

Postman + Charles in one tool is the right framing — the seam between "design a request" and "intercept what production is actually doing" is where most of us lose hours. Curious how mocking handles env-scoped rules (staging vs prod headers, per-branch base URLs) — that's where I usually end up writing glue. Open-source is a strong wedge here.