Чат с LLM из командной строки через OmniRoute

★ 7.9 · ai-tooling

cli-chat is a Claude Code skill that lets you send chat completions, stream responses, and run an interactive REPL session from the command line using OmniRoute. It provides three subcommands: `chat` for single-turn requests, `stream` for real-time token streaming with optional output saving, and `repl` for a persistent conversational session with history and resume support. All subcommands share a rich set of flags — model selection (`-m`), system prompt (`-s`), combo routing (`--combo`), and generation controls like `--temperature`, `--max-tokens`, and `--reasoning-effort`. It is well suited for developers and AI engineers who need to test LLM prompts in scripts, compare providers on the fly, or integrate language model calls into CLI workflows without writing boilerplate code.