automate-this

The `automate-this` skill is designed to transform repetitive manual tasks into automated workflows by analyzing screen recordings. Users provide a video of themselves performing a process, and the skill identifies the steps, applications, and actions involved. By converting visual and audio cues into structured automation scripts, it reduces time spent on tedious tasks and minimizes human error. This approach is particularly useful for professionals who frequently repeat the same actions across software applications, data entry, or file management.

Key capabilities include extracting frames from screen recordings at configurable intervals to balance detail and context limits, and optionally transcribing audio narration using Whisper if present. The skill can identify applications in use, sequential actions, data flow between steps, decision points, and repetitive patterns. It produces automation scripts that are compatible with the tools already installed on the user's machine, providing multiple complexity levels depending on user preference. Additionally, it performs prerequisite checks for essential tools such as ffmpeg and Whisper to ensure a smooth workflow.

This skill is ideal for individuals and teams looking to streamline repetitive processes, such as software developers, data analysts, administrative professionals, and power users managing complex workflows. It is particularly beneficial in environments where manual repetition is costly or error-prone. Users who record tutorials, repetitive reporting tasks, or multi-step operations across multiple applications can leverage `automate-this` to quickly generate reliable automation scripts and save significant time and effort.