Mise, IWE, and Nushell together give me a workflow that works every day.
The biggest productivity problem I know is not a lack of tools. It's the transitions between them.
You know what you need to do. You open the terminal. And then you spend five minutes trying to remember which folder you were in, what happened last, which note was relevant, and what you actually decided on Friday. Before you've even started, you're already a bit tired.
I've solved it with a combination of three things: mise as a task runner, IWE as a knowledge base, and Nushell as the glue that binds them together. Together, they form a workflow that starts with one command and doesn't require me to remember anything. So I can focus on the task!
Mise is a modern development tool that handles everything from language versions to environment variables—but what I use it for here is the task function. A .mise.toml file in the project folder defines all the commands I use regularly, with precise descriptions and logic. Instead of remembering iwe retrieve -k journal/$(date -v-1d +%Y-%m-%d) -d 1, I type mise run start. Instead of constructing a Nushell query from memory, I type mise run health. Mise isn't advanced—it's just a place where all commands live, so you don't have to reinvent them over and over.
The morning routine looks like this. The first command is mise run start, which fetches yesterday's journal via IWE and shows what I was working on. Not as a to-do list—as context. What did I decide? What did I stop in the middle of? What was waiting for something else? It takes ten seconds and replaces the mental energy that would otherwise be spent recreating the picture from scratch. Then, if there's no journal from today, mise run journal, which creates it and is ready for ongoing notes.
The weekly routine is mise run health. It runs two Nushell queries against iwe stats: one that finds orphaned notes—notes with content but no connections in or out of the graph—and one that finds notes that are highly referenced but very thin. These are two different types of knowledge base problems. Orphaned notes exist but don't function because no one finds them. Thin but popular notes are bottlenecks—everything points to them, but they don't deliver enough. Both are discovered in under a minute, and both can be addressed immediately.
This is where Nushell makes a real difference. A traditional shell would give me raw text back from iwe stats, which I would have to parse and interpret myself. Nushell treats the output as a structured table that can be sorted, filtered, and queried just like a database. The combination of iwe stats --format csv | from csv | sort-by total_incoming_refs --reverse | where words < 300 is not a trick—it's data analysis directly in the terminal, without opening a spreadsheet or writing a script.
What binds it all together is that mise, IWE, and Nushell all operate on the same premise: output is data, not text. IWE exports structured CSV. Nushell interprets it as a table. Mise stores the commands that connect them, so they don't need to be reinvented. And Claude Code can use it all—read the knowledge base via iwe retrieve, analyze it via Nushell, and create new notes via iwe new—because everything is built to be used programmatically.
The result isn't a fancy system. It's a system that lasts because it doesn't demand anything from me. I don't have to remember syntax, choose which folder I'm in, or restart the context from scratch. I type mise run start and I'm already underway.
That's what all good infrastructure should do.
We help companies think systematically about knowledge, AI, and the workflows that actually last in everyday life. If this sounds like something you need—book a no-obligation chat.