Prerequisite: first, make sure you have the Readwise MCP server or CLI installed (whichever you prefer!)
If you're using an AI agent, you can probably just tell your agent to install the skills from this repo and let it do the work:
Install the skills from here https://github.com/readwiseio/readwise-skills so that you can use them. If neither is installed yet, the Readwise MCP server is a remote MCP server hosted at https://mcp2.readwise.io/mcp OR the CLI can be installed via `npm i @readwise/cli`. You only need one of these installed to use the skills.
If that doesn't work:
npx skills is a very easy way to install these skills; it will set them to work with any LLM app you might use (Claude, Codex, Opencode, Cursor, etc). Run this from the terminal:
npx skills add readwiseio/readwise-skills
Copy the skills/ directory into your agent's skills path:
- Claude Code:
~/.claude/skills/or.claude/skills/in your project - Codex CLI:
~/.codex/skills/ - OpenCode:
~/.opencode/skills/
Installs the MCP server and all skills in one command:
/plugin marketplace add readwiseio/readwise-skills
/plugin install readwise@readwise-skills
In the Cowork sidebar: Customize → "+" → Add marketplace from GitHub → enter readwiseio/readwise-skills. Then browse and install the Readwise plugin.
We've started this repo off with a few skills we've really enjoyed using, but so much more is possible!
| Skill | Description |
|---|---|
| readwise-cli | Teaches your agent how to use the Readwise CLI — command syntax, setup, and common workflows |
| readwise-mcp | Teaches your agent how to use the Readwise MCP tools — tool names, parameters, and common workflows |
| triage | AI walks you through your inbox one article at a time, telling you what's worth your time and why |
| feed-catchup | Skim your Reader feed in batches — RSS, newsletters, Twitter digests — pull out the gems, mark the rest as seen |
| quiz | Test yourself on something you just read — graded like a smart colleague who also read the piece |
| book-review | Draft a long-form book review from your highlights — synthesizing the book with your broader reading history |
| reader-recap | Conversational briefing on your recent reading — what you finished, highlighted, and had to say about it |
| surprise-me | Analyze your reading history and tell you something surprising you don't know about yourself |
| highlight-graph | Visualize your highlights as an interactive 2D graph showing connections between ideas across sources |
| now-reading-page | Generate a "What I'm Reading" webpage from your library — host it on your personal site |
| build-persona | Build a reading profile from your highlights, tags, and history — powers personalization across all other skills |
Run the build-persona skill first to generate a reader_persona.md file. The other skills (triage, feed-catchup, quiz) read this file to personalize their output to your interests, goals, and reading style.
These skills are just a starting point, quickly hacked up by the Readwise team based on what's been for us to try. There is still so so much more that can be done. Please make a pull request adding your own favorite skills, or shoot us an email at any time :)