PROVIDER Β· the modern shell that treats data as tables
Nushell
Nushell applies the everything-is-data idea in Rust, turning command output into typed tables where where, select, and group-by handle JSON and CSV far more clearly than awk as a modern cross-platform terminal.
Pros
Cons
Product Information
Detailed Notes
Overview
Nushell was open sourced in 2019 and written in Rust around the idea that everything is data, where command output is a table with named typed columns that can be filtered, sorted, and grouped like SQL before exporting via to json or to yaml.
For Programming Languages & Frameworks work the value is data exploration with cross-platform consistency, since JSON, CSV, and logs need no awk wizardry when where, select, and group-by apply directly, the same commands behave identically on Windows, macOS, and Linux, and clear errors feel closer to a modern language than bash.
See the current state, because syntax is wholly incompatible with bash and POSIX so existing scripts will not run, the ecosystem is early with some commands and plugins unstable, rapid releases occasionally break configuration and syntax, and servers do not install it by default limiting ops adoption.
Core Strengths
- Typed tabular output makes filtering, sorting, and grouping feel like SQL
- Native parsing and conversion for JSON, YAML, CSV, and TOML
- Written in Rust with identical behaviour on Windows, macOS, and Linux
- Clear error messages with modern language-like syntax
Product Ecosystem
Data pipelines
where, select, sort-by, group-by, and each operate on tabular data.
Format conversion
from json, yaml, csv, and toml read in while to json, yaml, csv write out for interoperability.
Configuration
config nu manages settings with custom commands and modular scripts.
Drawbacks
- Fully incompatible with bash and POSIX so existing scripts cannot be reused
- An early ecosystem with some unstable commands and plugins
- Rapid releases occasionally break configuration and syntax
- Not installed on servers by default, raising adoption cost
Use Cases
- JSON and CSV exploration (β β β β β ) - Clear tabular handling
- Cross-platform personal terminals (β β β β ) - Consistent on three platforms
- Replacing some awk use cases (β β β β ) - Friendlier syntax
- Production ops scripts (β ) - Ecosystem and adoption gaps
Pricing
| Plan | Price | Scope |
|---|---|---|
| Nushell and source | $0 | MIT licensed, free |
| Plugin ecosystem | $0 | Open source and free |
| Commercial support | None official | Community maintained |
Entirely free with cost only in learning new syntax, fitting personal data exploration today rather than replacing production bash operations scripts.
FAQ
-
Can Nushell run bash scripts? No, the syntax model differs entirely, though it can call scripts through bash -c and is not a POSIX shell.
-
Does it replace awk? It is far clearer for structured JSON and CSV work while awk stays lighter for plain text streams.
-
Is it production ready? Fine for personal exploration and internal tools, with critical ops scripts best migrated later.
Code Examples
Structured data pipelines
open plans.json
| where price < 80
| sort-by price
| select name price
| to md
open parses JSON, CSV, or TOML automatically and output renders as a table, Nushell's core selling point.
Custom commands
def "check hosts" [] {
$in | each { |h| { host: $h, status: (http get $"https://($h)/health" | get status) } }
}
["hk-1" "us-1"] | check hosts
Community & Resources
Communities and official channels
- Nushell Site β Features and installation
- GitHub nushell/nushell β Source, issues, and releases
- Nushell Discord β Official community chat
Learning resources and tools
- Nushell Book β The official book from basics up
- Nushell Cookbook β Recipes by task
- Awesome Nu β Plugins and configurations