The free brand voice analyzer: paste your website, hear your own voice back
There's now a free tool at borker.xyz/free/brand-voice that reads your website and plays your brand voice back to you, complete with voice sliders, personality traits, and sample posts. No account needed. This release also makes auto-publish safer by default (no sensitivity rules means content holds for review) and stops onboarding from building a brand profile off a page it couldn't read.
Three things this time: a free tool anyone can try without signing up, a safer default for auto-publish, and an onboarding fix for the "why does my brand profile say Unknown" case.
Free brand voice analyzer
Borker's whole pitch is that it learns your voice and writes like you. That's also the part people rightly don't take on faith. So we put the first step of the product on a public page: borker.xyz/free/brand-voice.
Paste your website URL. Borker reads the page, models the voice it finds, and plays it back to you: where you sit between casual and formal, neutral and warm, verbose and concise, plus the personality traits it picked up, each with a concrete justification pulled from your actual writing. Then it writes sample posts in that voice.
It takes about a minute, shows you its work while it runs, and asks for nothing up front. At the end you pick: start a trial with this analysis as your ready-made brand config, or have the full report emailed to you. Both paths are fine with us. The analyzer is the honest demo, and if the profile it writes doesn't sound like you, we'd rather you find out on the free page than three days into a trial.
The screenshots show Borker analyzing borker.xyz. We use our own tools around here.
What else shipped
Auto-publish holds everything until you set a sensitivity rule
Auto-publish mode ships generated content without review. Sensitivity rules (keywords and natural-language principles you configure under Settings → Schedule) are the guardrail: anything that matches a rule is held for your review instead of published.
The gap was the empty state. A workspace on auto-publish with zero rules configured had no guardrail at all, and everything shipped straight to your channels. That's the riskiest possible configuration behaving like the most confident one.
The default is now flipped. On auto-publish with no sensitivity rules, every generated item is held as a draft, marked with the reason, and waits in your Pipeline's review tab. Add at least one keyword or principle and autopilot resumes for everything that doesn't match it. The "Suggest with AI" button will draft a starter rule set from your brand profile if you don't want to start from a blank list.
Onboarding checks that it can read your website
During onboarding, Borker reads your website to build your brand profile. But some URLs don't lead anywhere readable: a LinkedIn page behind a login wall, a site that blocks bots, a typo'd domain. Previously the analysis would "succeed" anyway and quietly save a brand profile full of placeholders, with a company name of "Unknown" and a mission statement that said, in so many words, nothing.
Now Borker checks what it could read before building anything. If there isn't enough real content, you get a plain explanation and two ways forward: fix the URL, or add a short description of your brand and analyze again. Nothing gets saved until the analysis has something real to work with.
Bug fixes
- Long posts to Farcaster now publish correctly instead of failing on length.
- Disconnected channels announce themselves. If a channel's connection expires, you now get a reconnect prompt instead of silent publish failures.
New here? Borker is the AI content engine for founders.
We learn your voice, watch your news feeds, and ship posts to X, LinkedIn, Farcaster and your blog while you build the actual product.