AI NewsOpen sourceAnnouncement
new-term-radar catches new AI tool names before keyword tools list them
new-term-radar is a free MIT-licensed daily job that reads X, Hugging Face, GitHub and sitemaps and prints a ranked list of new AI tool names before most keyword research tools list them. It has 298 stars in seven days.
Image: GitHub
Why it mattersContent and SEO teams who want to rank or get cited for the name of a new AI tool only win if they publish the week it is coined, and this script tells them which name to pick.
Keyword research tools name a new AI tool roughly a week after a human on X has already named it, and a new Python script tries to close that gap for anyone writing about AI tools. new-term-radar, posted by zhuyansen under the MIT licence, is a daily job that reads five sources and prints a ranked list of candidate names, with a Google Trends check attached. The repository has 298 stars in seven days.
The sources are X posts from a configured list of accounts, Hugging Face trending models, GitHub repositories gaining more than 200 stars in a day, sitemaps of competitor sites and newly registered domains that are already ranking. A sixth optional signal reads the user's own Google Search Console for queries they are already picking up.
What it is for
The target audience the README names is "people who make pages, posts or products around new terms: SEO for AI tool sites, newsletters, directories". A new AI tool's name is a short-lived search phrase: it reaches peak volume the week it is announced, and most of the pages that end up ranking for it were posted in the first few days. A content team that waits for a keyword volume number from a research tool is already late.
The output is a Markdown table and a JSON file. Each candidate carries a name score, Google Trends verification and a list of search variants. The README's own sample lists "gridcraft" and "artex" as detected breakouts, each with a star-velocity figure next to it. gridcraft is a storytold.com Rust reimplementation of Excel the trending radar has already picked up. The list is of words that have started being written about, regardless of whether each one maps to a product.
How it runs
The project is extracted from a radar that runs daily behind agentskillshub.top, and the README ships two workflow files for GitHub Actions. The sources command runs every eight hours and refreshes the raw signals, and the daily command runs once a day and writes a report. A local run uses a config.json file and plain Python.
Costs are stated plainly. Hugging Face and GitHub are free. The X signal uses the AIsa API at $0.003 per account per sweep, which the README puts at about $10 a month for a configured list. The keyword table uses another paid API that works out to roughly $5 a month, for a total of about $15.
A reader whose job is to write the first good page for a new AI tool has to know its name on day one. new-term-radar is one way to find out, and the ceiling of what it can replace is a weekly keyword research export that already lists everything a day late. The limit is noise: not every new term reaches a stable search level, and running the script costs an afternoon of configuration before the first useful report.
Source
This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.
Get AI News in your inbox
New developer tools, model and agent releases, and how teams are actually using them to release software. Short, and only when there is something worth reading.
