Show HN: Promptscout a local prompt enricher for Claude Code (github.com)

Promptscout, a new command-line interface (CLI) tool, has been released to enhance coding agent prompts by enriching them with contextual information from a local codebase. Designed as a plugin for Claude Code, this tool operates entirely on the user's machine without requiring API keys or cloud access. It utilizes a local language model (LLM), specifically Qwen 3 4B, to analyze the user's prompt and scan the codebase for relevant files and snippets using tools like ripgrep and git. By appending this context to the original prompt, coding agents can directly dive into relevant work, significantly improving efficiency and reducing the cost associated with generating unnecessary codebase awareness. The significance of Promptscout lies in its ability to streamline the development workflow. By pre-fetching contextual data, it allows programmers to receive immediate, focused responses from coding agents like Claude Code, Cursor, or Copilot. This not only saves time but also cuts down on token usage, as coding agents will spend less time figuring out which files are pertinent to the task at hand. As coding projects grow in complexity, tools like Promptscout can enhance productivity and improve the overall coding experience by providing precise, contextually rich support.