wtf is akta.pro-mcp-skills?

This repository holds skill files that teach AI agents how and when to use akta.pro's data about private companies and related news. A skill file is a set of instructions that tells an agent what kind of information is available and when it makes sense to reach for it, while a separate connection technology called MCP is what actually lets the agent pull that data in. Because MCP is a shared standard, the same connection works across many different AI tools, including Claude, ChatGPT, Claude Code, Cursor, VS Code, and OpenCode. Akta.pro itself is a paid data service covering more than 20 million private companies, each described by over 70 structured fields such as legal name, founding year, headcount, revenue estimates, funding history, technology stack, and ownership relationships. On top of that company database sits a news and signals layer, which tracks companies, whole industries, or custom topics defined in plain language, and returns deduplicated, entity resolved news items with an AI generated summary, sentiment label, and impact score attached to each one. To use it, you first create a free akta.pro account, which comes with a starting allowance of 25 credits to try out company lookups and news signals before paying for anything. From the akta.pro dashboard you then connect your AI tool of choice using a single shared server address, following slightly different setup steps depending on whether you are configuring Claude, ChatGPT, Claude Code, Cursor, VS Code, or OpenCode, most of which involve pasting the server address and an API key into a settings file or menu. Once the connection is set up, you place the relevant skill file from this repository wherever your particular agent expects to find skills. From that point on, asking the agent a company research question should automatically trigger the skill and let it pull structured company data and relevant news directly from akta.pro. The repository itself is organized simply, containing this README, a folder of skill files, and a folder of sample outputs showing what a competitive intelligence query looks like in practice.