hardhackerlabs/podwise-cli

What it solves

Podcasts are long, but their value is often buried in transcripts. This project lets you skip the hours of listening by turning any episode into structured insights — summaries, transcripts, Q&A, mind maps, highlights, and more — without needing a browser or human in the loop. It solves the problem of making podcast content accessible to AI agents and automation pipelines.

How it works

The tool is a CLI client for podwise.ai. After installing and authenticating (the CLI saves your API key), you can run commands to fetch trending episodes, ask questions answered from transcripts, search by keyword, process new episodes from URLs or local files, and retrieve results like summaries or transcripts. It also includes a built-in MCP server (started with podwise mcp) that exposes the same capabilities as standardized tools — such as search_episode, get_summary, process, and ask — so AI assistants like Claude Desktop, Gemini CLI, or Cursor can call them directly. There are also pre-built agent skills (workflows) for tasks like weekly recaps, topic research, or language-learning flashcards.

Who it's for

Developers and users who want to integrate podcast knowledge into AI agents, skills runtimes, or automation pipelines — for example, building a personal research assistant that can answer questions based on podcast content, or a system that automatically generates notes from your favorite shows.

Highlights

• Fetch structured insights (summaries, transcripts, Q&A, mind maps, highlights) from any podcast episode via simple CLI commands.
• Works with multiple input sources: Podwise URLs, Xiaoyuzhou, YouTube, or local audio files.
• Includes an MCP server for direct integration with AI assistants, exposing 19+ tools like ask, search_episode, and get_transcript.
• Pre-built agent skills with workflows for catching up, weekly recaps, episode debates, language learning (Anki flashcards), and more.
• Easy installation via Homebrew, a curl script, or building from source.