Why I Built Brandomica: I Launched an Unsafe Brand Name

The failure

I had a name I liked. Domain was available. Twitter handle was free. I got excited and started building. The trademark search happened almost as an afterthought, weeks later. Active registration. My class. Everything I’d done was useless.

The name was “available” in every obvious sense. But it wasn’t safe — there was a legal risk I didn’t see because I checked the wrong things first. This is the core problem that brand name safety assessment solves: evaluating risk across multiple dimensions, not just checking individual availability signals.

The second attempt

I picked a new name and started with trademarks this time. Clear. Domains — .com taken by a parked page, but .io was free. Social handles — Twitter existed but hadn’t posted in years. Then I Googled the name. A company in Brazil was using it in the same space.

Not a legal threat, but a collision risk. Anyone Googling my product would find their company first. The name was technically available but not safe to build on — a distinction that no individual availability check could surface on its own.

The fragmented workflow

I spent 30+ hours across three candidate names. The time was painful, but the real problem was that no tool gave me an integrated safety assessment. Each tool checked one thing: a registrar checked domains, a trademark database checked marks, a social platform checked handles. None connected the dots.

A domain registrar that also checks Twitter doesn’t evaluate trademark risk. A trademark search doesn’t flag that a name means something offensive in German. A Google search doesn’t measure phonetic similarity to existing marks. I was assembling partial signals from a dozen tools and making safety judgments by gut feeling.

What was missing

Not more availability checks. A safety assessment — a single evaluation that considers trademarks, domains, social handles, web presence, linguistics, and phonetics together, and produces a score that reflects how much risk a name actually carries.

A taken TikTok handle is not the same severity as an active trademark. A tool that treats them the same (green dot / red dot) isn’t assessing risk — it’s listing results. What I needed was weighted risk signals and explicit blockers that flag the dangers that should genuinely stop you from using a name.

Building Brandomica

Brandomica checks 19 sources and produces a 0–100 safety score across six risk signals: legal, collision, impersonation, linguistic, phonetic, and coverage. It flags blockers and recommends specific actions. Domains come with real pricing (Year 1, renewal, 3Y TCO). Trademark checks use live registry data with confidence and missing-critical-checks reporting.

The shift to agent-first

After launching the web app, I realized the manual workflow was already outdated. In 2026, AI agents generate brand name candidates on their own. What they lack is external data for safety verification — live trademark databases, domain registrars, social platforms, search engines.

So the tool grew into four interfaces: MCP server for AI agents calling safety tools natively, CLI for terminal and CI/CD pipelines, REST API for custom integrations, and the web app for visual exploration. The real workflow has moved to agents: generate 10 names, batch-check safety, filter by score, present only the safe ones. See One Tool, Four Interfaces for details.

Free, no account required

Brand safety shouldn’t be locked behind a signup wall. Every interface is free — npx brandomica-mcp-server for agents, npx brandomica check <name> for the terminal. If the story above sounds familiar, point an agent at your next name. It usually takes a few seconds to get a safety result, depending on provider latency.