Set up shell completion for the Elastic CLI

elastic completion <shell> prints a wrapper script that hooks the CLI into your shell's tab-completion system. Bash, Zsh, Fish, and PowerShell are supported.

The wrapper is dynamic: each <tab> shells out to elastic to ask which candidates apply, so completions stay in sync with the installed CLI version and include context names from your config file.

Bash

		# System-wide (writable by root):
elastic completion bash | sudo tee /etc/bash_completion.d/elastic > /dev/null

# Per-user (no sudo required):
mkdir -p ~/.local/share/bash-completion/completions
elastic completion bash > ~/.local/share/bash-completion/completions/elastic
		
	

Then open a new shell.

Zsh

		# Drop the script into the first directory in $fpath:
elastic completion zsh > "${fpath[1]}/_elastic"

# Make sure compinit is enabled in ~/.zshrc:
autoload -Uz compinit && compinit

Or, for a one-shot install in the current session:

eval "$(elastic completion zsh)"

Fish

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

Fish picks up new completion files automatically.

PowerShell

elastic completion powershell | Out-File -Encoding utf8 -Append $PROFILE

Then open a new PowerShell session.

What gets completed

Top-level commands and groups (stack, cloud, docs, config, …)
Nested subcommands at every depth
Long option flags (--json, --use-context, command-specific options)
Context names for --use-context (read from the active config file)
The es / kb aliases work identically to the canonical stack es / stack kb forms

Completion respects the active commands.allowed / commands.blocked policy: commands you have restricted yourself out of do not appear as candidates.