What problem does it solve? Manually uploading code to remote Linux servers and starting Python consumers on each machine is repetitive and error-prone. This Skill automates function-level remote deployment for funboost: it uploads your project over SFTP, kills old consumer processes, and starts the specified @boost function with multi-process consumption on the target machine. ## Core Features & Use Cases - Function-level deployment: Deploy a single @boost function (not a whole script) to a specific machine via my_task.fabric_deploy(host, port, user, password, process_num=2). - Automated upload and process management: Syncs project code over SFTP with exclusion rules and incremental upload, then kills old processes by a unique funboost mark before starting new ones. - Production persistence options: Configure pty=False so remote consumers keep running after the local deploy script exits, and use SSH private keys via pkey_file_path. - Use Case: You have two Linux servers and want task_a running with 2 processes on server 1 and task_b with 1 process on server 2, all deployed from one local Python script without logging into either machine. ## Quick Start Ask the AI to deploy your @boost decorated function to a remote Linux server by calling its fabric_deploy method with the host, port, SSH credentials, and desired process count.
