What is local-client-prospector-skill?

Local Client Prospector is a Codex skill, meaning it is an add-on for the Codex AI assistant, built to help freelancers, agencies, or small business consultants find potential clients. Instead of searching manually for local shops, gyms, restaurants, salons, or clinics that might need a website or a better online presence, this tool automates the search using browser-assisted research. Once installed, it looks up nearby businesses and checks whether each one already has a proper standalone website or only relies on social media pages. Based on what it finds, it assigns each business a lead score of Hot, Warm, Low, or Skip, giving the user a quick sense of which prospects are worth contacting first. The results come back as a table readable directly in chat, or in CSV-style rows that are easy to paste into a spreadsheet. It also works alongside Codex's spreadsheet skill, so results can be exported straight into an XLSX file with columns for phone number, website, social links, source, and notes. Getting started is simple: running one command, npx local-client-prospector-skill, installs the skill into the Codex skills folder, and it becomes available the next time Codex starts. From there, a user can type a plain language request, such as asking for gyms and shops within a set distance of a specific location, and the skill returns a formatted lead list, replying in whatever language the request was written in. The project describes itself as an assisted research tool rather than a scraper meant for pulling large volumes of data automatically. It is built to check public information responsibly and cross-reference business details rather than bypass any access restrictions, which suggests it is aimed at small-scale, targeted prospecting rather than mass data collection. This could be useful for freelance web designers, marketing consultants, or small agencies looking for local businesses that lack a proper online presence and might want help building one. The README is fairly short and focused mainly on installation and usage instructions, without much detail on the underlying technical implementation.