COMMAND LINE

Your data.
Your terminal.

Upload files, query datasets, stream logs. All from the command line. Single binary, instant startup, works everywhere.

TERMINAL

$ lite upload products.csv --name products ✓ Dataset created Name products Rows 2,847 Columns id, name, category, price, rating API https://api.liteio.dev/d/products $ lite q products category=electronics --sort -price --limit 3 NAME PRICE RATING Pro Headphones 299.00 4.9 USB-C Hub 79.99 4.7 Wireless Mouse 59.99 4.6

INSTALL

One line. Any platform.

Single binary, no runtime dependencies. Built in Go for instant startup.

CURL

curl -fsSL https://api.liteio.dev/install.sh | sh

macOS & Linux

HOMEBREW

brew install liteio/tap/lite

macOS & Linux

NPM

npm install -g @liteio/cli

Any platform with Node.js

go install github.com/go-mizu/lite@latest

Any platform with Go

QUICKSTART

Three commands to a live API.

Upload

Drop any CSV, JSON, TSV, or Excel file. Schema detected automatically.

UPLOAD

$ lite upload sales.json --name sales Uploading sales.json (89 KB)... ✓ Dataset created Name sales Rows 12,450 Columns order_id (integer), customer (text), product (text), amount (number), date (date), region (text) API https://api.liteio.dev/d/sales

Query

Filter, sort, search, paginate. All from the command line.

QUERY

$ lite q sales region=west amount:gt=500 --sort -amount --limit 5 CUSTOMER PRODUCT AMOUNT DATE Acme Corp Enterprise 4,200 2026-03-20 DataTech Inc Pro License 2,800 2026-03-19 CloudBase Team Plan 1,500 2026-03-18 NetOps Ltd Enterprise 1,200 2026-03-17 DevHub Co Pro License 890 2026-03-15 Showing 5 of 1,847 results.

Open in the browser, pipe to other tools, or give the URL to your AI.

$ lite open sales Opening https://api.liteio.dev/d/sales in browser... $ lite export sales --filter region=west --format csv | head -3 order_id,customer,product,amount,date,region 8291,Acme Corp,Enterprise,4200,2026-03-20,west 8274,DataTech Inc,Pro License,2800,2026-03-19,west

COMMANDS

Everything you need. Nothing you don't.

Short aliases for speed. --json on every command for scripts.

DATA

upload <file> --nameUpload file, get APIup

query <name>Filter, sort, searchq

schema <name>Show column types

stats <name>Column statistics

get <name> <row>Get single rowcat

export <name>Export CSV/JSON/JSONL

MANAGE

lsList datasets

push <name> <file>Replace dataset

append <name> <file>Add rows

delete <name>Delete datasetrm

open <name>Open in browser

logs [dataset]Stream access logs

ACCOUNT

loginAuthenticate

logoutClear credentials

whoamiShow identity

config [key] [val]Get/set config

completion <shell>Shell completions

versionPrint version

QUERY ENGINE

Nine operators. Any combination.

Filter by any column with typed operators. Chain multiple filters with AND. Sort by any field, ascending or descending. Full-text search across all text columns.

Positional shorthand lite q products price:gt=100

Multi-sort --sort -price,name for multi-field

Full-text search --search "wireless headphones"

field=valueexact match

field:gt=100greater than

field:gte=100greater or equal

field:lt=50less than

field:lte=50less or equal

field:ne=booksnot equal

field:like=wire%pattern match

field:in=a,b,cin set

field:null=trueis null

OUTPUT

Human. Machine. You choose.

Colored tables for humans. JSON for scripts. Quiet mode for pipes. Auto-detects when output is piped and switches to machine-friendly format.

--json on every command, with field selection

--quiet minimal output (IDs, counts only)

Auto-detect no color or spinners when piped

TABLE (default)

$ lite ls NAME ROWS COLS CREATED products 2,847 6 2h ago sales 12,450 14 1d ago users 891 4 3d ago

JSON (--json)

$ lite ls --json --fields name,rows [{"name":"products", "rows":2847}, {"name":"sales", "rows":12450}, {"name":"users", "rows":891}]

OBSERVABILITY

Watch your API live.

Stream access logs in real time. See who's querying your data, what they're filtering, and how fast it responds. Filter by status code or time range.

Real-time streaming logs appear as requests come in

Filter by status --status 4xx for errors only

Time range --since 1h for recent history

LOGS

$ lite logs products Streaming logs... (Ctrl+C to stop) 12:03:41 GET ?category=electronics&sort=-price 200 23ms 12:03:44 GET /schema 200 8ms 12:04:02 GET ?q=wireless 200 31ms 12:04:15 GET ?format=csv 200 89ms 12:04:28 GET ?page=2&limit=50 200 19ms 12:05:01 GET ?price:gt=1000 200 12ms 12:05:33 GET /stats 200 45ms

SCRIPTING

Pipes. Scripts. CI/CD.

Every command speaks JSON. Meaningful exit codes. Pipe-friendly. Capture dataset IDs, chain queries, integrate with jq, Python, or your CI pipeline.

Exit codes 0 success, 1 error, 2 usage, 3 auth, 4 not found

LITE_TOKEN env var for CI authentication

--quiet just the ID or count, for capture

SCRIPT

#!/bin/bash # Upload with a name lite upload data.csv --name products # Query by name, pipe to jq lite query products --json | jq '.data | length' # Upsert: update if exists, create if not if lite schema products --quiet 2>/dev/null; then lite push products data.csv else lite upload data.csv --name products fi

GITHUB ACTIONS

- name: Upload dataset env: LITE_TOKEN: ${{ secrets.LITE_TOKEN }} run: | curl -fsSL api.liteio.dev/install.sh | sh lite upload ./data/products.csv --name products

CONFIGURATION

Zero config to start. Full control when you need it.

Auth Flow

Browser-based login. Token stored securely. Override with env var for CI.

1 lite login opens browser

2 Authenticate in browser

3 Token stored in ~/.config/lite/

AUTH PRIORITY

1 --token flag

2 LITE_TOKEN env var

3 Stored credential

4 No auth (public only)

Config Files

Global config in ~/.config/lite/. Optional project config in lite.toml.

~/.config/lite/config.toml

format = "table" endpoint = "https://api.liteio.dev" default_limit = 25 no_color = false

lite.toml (project)

[aliases] prod = "products" ord = "orders" # Use short aliases: # lite query prod price:gt=100 # lite push ord ./data/orders.csv

ENVIRONMENT

Environment variables.

LITE_TOKEN Auth token (overrides stored credential)

LITE_ENDPOINT API endpoint (for self-hosted)

LITE_FORMAT Default output format (table, json, csv)

LITE_NO_COLOR Disable color output

NO_COLOR Standard no-color convention

COMPLETIONS

Tab to complete.

Dynamic completions for dataset IDs, column names, config keys, and filter operators.

Bash

lite completion bash > /etc/bash_completion.d/lite

Zsh

lite completion zsh > "${fpath[1]}/_lite"

Fish

lite completion fish > ~/.config/fish/completions/lite.fish

ERRORS

Errors that help.

Every error tells you what went wrong, why, and how to fix it. Exit codes are meaningful for scripts. No stack traces, no guessing.

0 Success

1 General error

2 Usage error

3 Auth error

4 Not found

5 Rate limited

ERROR

$ lite query orders Error: Dataset not found No dataset named "orders". Run lite ls to see your datasets.

ERROR

$ lite upload huge.csv Error: File too large (24 MB) Maximum file size is 10 MB. Split the file or compress it.

Install in 5 seconds. Upload in 10.

One line to install. One command to turn any file into a live API.

curl -fsSL https://api.liteio.dev/install.sh | sh

API Reference Learn more

# CLI lite: upload files, query data, manage datasets from your terminal. Single binary. No runtime dependencies. Works without login for public datasets. ## Install

bash # macOS / Linux curl -fsSL https://api.liteio.dev/install.sh | sh # Homebrew brew install liteio/tap/lite # npm npm install -g @liteio/cli # Go go install github.com/go-mizu/lite@latest

## Quick Start

bash # Upload a file (pick a name, it becomes your endpoint) lite upload products.csv --name products # Query by name with filters and sort lite query products --filter category=electronics --sort -price # Full-text search lite query products --search "wireless headphones" # Show schema lite schema products # Open in browser lite open products

## Authentication

bash # Login (opens browser) lite login # Check identity lite whoami # Logout lite logout

Auth token stored in ~/.config/lite/credentials.json. Override with LITE_TOKEN env var or --token flag. Priority: --token flag > LITE_TOKEN env var > stored credential > no auth (public only). ## Commands ### Upload

bash lite upload <file> --name <name> lite up <file> --name <name> # alias

FlagDescription

--name <name>Choose a name for your dataset (becomes the API path)

--privateRequire auth to read

--schema <file>JSON column type overrides

--delimiter <char>Force CSV delimiter

--no-headerFirst row is data, not headers

`--format csv\tsv\json\jsonl\xlsx`Force format

--jsonJSON output

--quietPrint only dataset ID

Supported formats: CSV, TSV, JSON, JSON Lines, Excel (.xlsx). Auto-detected from extension. ### List Datasets

bash lite ls

FlagDescription

--jsonJSON output

--quietPrint only IDs

--fields <cols>Select JSON fields

### Query

bash lite query <name> [filters...] lite q <name> [filters...] # alias

FlagDescription

--filter <field:op=val>Filter (repeatable). Ops: gt, gte, lt, lte, ne, like, in, null

--sort <fields>Sort, prefix - for desc. Comma-separated

--search <query>Full-text search across text columns

--limit <n>Max rows (default 25)

--page <n>Page number

--offset <n>Skip rows

--columns <cols>Select columns

`--format table\json\csv\jsonl`Output format

--jsonAlias for --format json

--no-headerOmit table header

--wideDon't truncate columns

Filter shorthand:

bash # Positional filters after dataset name lite q products category=electronics price:gt=50 rating:gte=4

### Schema

bash lite schema <name>

Shows column names, types, nullability, and row count. ### Stats

bash lite stats <name>

Shows min, max, avg, null count, unique count, and top values per column. ### Get Row

bash lite get <name> <row_number> lite cat <name> <row_number> # alias

### Delete

bash lite delete <name> lite rm <name> # alias

Requires confirmation. Use --yes to skip in scripts. ### Push (Replace)

bash lite push <name> <file>

Replace entire dataset. Same ID, same URL. ### Append

bash lite append <name> <file>

Add rows to existing dataset. ### Export

bash lite export <name> [flags]

FlagDescription

`--format csv\json\jsonl`Export format (default: csv)

-o <file>Output file (default: stdout)

--filter, --sort, --searchApply before export

### Logs

bash lite logs [dataset]

Stream access logs in real time.

FlagDescription

--since <duration>Show logs from (e.g., 1h, 30m)

--status <code>Filter by status (e.g., 4xx, 200)

--jsonJSON output

--no-followDon't stream, print and exit

### Open

bash lite open <name>

Open dataset API URL in default browser. ## Output Formats Default: Human-readable tables with color. Auto-sized to terminal width. --json: Machine-readable JSON on every command. Supports --fields for field selection. --quiet: Minimal output (just IDs or counts). For scripting. Pipe detection: When stdout is not a TTY, color and spinners are disabled automatically. Query output defaults to JSON when piped. ## Configuration ### Config File ~/.config/lite/config.toml

toml format = "table" endpoint = "https://api.liteio.dev" default_limit = 25 no_color = false editor = "vim"

### Commands

bash lite config # Show all lite config <key> # Get value lite config <key> <value> # Set value lite config --reset # Reset to defaults

### Project Config Optional lite.toml in project directory:

toml [aliases] prod = "products" ord = "orders" # Use short aliases: # lite query prod --filter price:gt=100 # lite push ord ./data/orders.csv

### Environment Variables

VariableDescription

LITE_TOKENAuth token

LITE_ENDPOINTAPI endpoint

LITE_FORMATDefault output format

LITE_NO_COLORDisable color

NO_COLORStandard no-color convention

## Shell Completions

bash # Bash lite completion bash > /etc/bash_completion.d/lite # Zsh lite completion zsh > "${fpath[1]}/_lite" # Fish lite completion fish > ~/.config/fish/completions/lite.fish

Dynamic completions for dataset IDs, column names, and config keys. ## CI/CD

yaml # GitHub Actions - name: Upload dataset env: LITE_TOKEN: ${{ secrets.LITE_TOKEN }} run: | curl -fsSL https://api.liteio.dev/install.sh | sh lite upload ./data/products.csv --name products

Scripting patterns:

bash # Upload with a name lite upload data.csv --name products # Query by name, pipe to jq lite query products --json | jq '.data | length' # Upsert: update if exists, create if not if lite schema products --quiet 2>/dev/null; then lite push products data.csv else lite upload data.csv --name products fi # Export for another tool lite export products --filter status=active --format csv | python analyze.py

## Exit Codes

CodeMeaning

0Success

1General error

2Usage error (bad arguments)

3Auth error

4Not found

5Rate limited

## Links - API Reference: Full HTTP API - Source: CLI source code - Releases: Download binaries

Your data.Your terminal.

One line. Any platform.

Three commands to a live API.

Upload

Query

Share

Everything you need. Nothing you don't.

Nine operators. Any combination.

Human. Machine. You choose.

Watch your API live.

Pipes. Scripts. CI/CD.

Zero config to start. Full control when you need it.

Auth Flow

Config Files

Environment variables.

Tab to complete.

Errors that help.

Install in 5 seconds. Upload in 10.

Your data.
Your terminal.