src/bx_skills/catalog_skills/bx-bash-reference/SKILL.md
Use when writing, reviewing, or debugging Bash scripts, when needing exact syntax for shell constructs, parameter expansions, redirections, builtins, test expressions, arrays, or any Bash 5.3 feature. Use when unsure about quoting rules, expansion order, conditional operators, trap behavior, shopt options, or specific builtin options and arguments.
npx skillsauth add bitranox/bx_skills bx-bash-referenceInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Complete reference for GNU Bash 5.3 (May 2025). Covers all shell syntax, builtins, variables, expansions, redirections, and features. Use this skill whenever writing or reviewing bash scripts to ensure correctness.
This file is a scannable hub. Each section below summarizes key constructs with compact tables. Use the Read tool to load referenced files identified as relevant for full syntax, edge cases, and examples:
shell-syntax-and-commands.mdfunctions-parameters-expansions.mdredirections-and-execution.mdset, shopt, declare, read, printf, trap, etc.): shell-builtins.mdBASH_REMATCH, PIPESTATUS, EPOCHSECONDS, etc.): shell-variables.mdbash-features.mdjob-control-readline-history.md${ command; }, GLOBSORT, BASH_MONOSECONDS)sh scripts (this covers Bash-specific features beyond POSIX)bash_clean_architecture instead)| File | Contents |
|--------------------------------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| shell-syntax-and-commands.md | Quoting, comments, reserved words, pipelines, lists, compound commands (if/for/while/case/select/[[/(( ), grouping, coprocesses |
| functions-parameters-expansions.md | Shell functions, positional/special parameters, ALL expansion types (brace, tilde, parameter, command substitution, arithmetic, process substitution, word splitting, globbing, pattern matching) |
| redirections-and-execution.md | All redirection types, here docs/strings, file descriptors, command search/execution, execution environment, exit status, signals, shell scripts |
| shell-builtins.md | All Bourne shell builtins, all Bash builtins, set options, shopt options, special builtins |
| shell-variables.md | All Bourne shell variables, all Bash variables (BASH_, COMP_, HIST*, READLINE_*, etc.) |
| bash-features.md | Invocation options, startup files, interactive shell behavior, conditional expressions, arithmetic, aliases, arrays, directory stack, prompt control, restricted shell, POSIX mode, compatibility mode |
| job-control-readline-history.md | Job control (bg/fg/jobs/disown), readline configuration, all bindable commands, vi mode, programmable completion (complete/compgen/compopt), history expansion |
| I need to... | Read |
|--------------------------------------------------------------------------------------|--------------------------------------|
| Write a function, use parameters/expansions, do string manipulation | functions-parameters-expansions.md |
| Use if/for/while/case/select/[[ ]]/(( )), or understand quoting | shell-syntax-and-commands.md |
| Redirect I/O, use here docs/strings, understand fd management | redirections-and-execution.md |
| Look up a builtin (set, shopt, declare, read, printf, trap, etc.) | shell-builtins.md |
| Check a shell variable (BASH_REMATCH, PIPESTATUS, EPOCHSECONDS, etc.) | shell-variables.md |
| Understand startup files, arrays, arithmetic, POSIX mode, or conditional expressions | bash-features.md |
| Work with job control, readline, completion, or history expansion | job-control-readline-history.md |
Full details:
shell-syntax-and-commands.md(section 1)
| Syntax | Behavior |
|-----------|------------------------------------------------------------------|
| \x | Escape single character |
| 'text' | Literal string, no expansion |
| "text" | Allows $, `, \, ! expansion |
| $'text' | ANSI-C escapes: \n, \t, \e, \xHH, \uHHHH, \UHHHHHHHH |
| $"text" | Locale-specific translation |
Full details:
functions-parameters-expansions.md(section 3.4)
| Syntax | Result |
|------------------------|----------------------------------|
| ${var:-default} | Use default if var unset/null |
| ${var:=default} | Assign default if var unset/null |
| ${var:+alternate} | Use alternate if var IS set |
| ${var:?error} | Error if var unset/null |
| ${#var} | String length |
| ${var:offset:length} | Substring |
| ${var#pattern} | Remove shortest prefix match |
| ${var##pattern} | Remove longest prefix match |
| ${var%pattern} | Remove shortest suffix match |
| ${var%%pattern} | Remove longest suffix match |
| ${var/pat/str} | Replace first match |
| ${var//pat/str} | Replace all matches |
| ${var/#pat/str} | Replace if matches beginning |
| ${var/%pat/str} | Replace if matches end |
| ${var^pattern} | Uppercase first char |
| ${var^^pattern} | Uppercase all chars |
| ${var,pattern} | Lowercase first char |
| ${var,,pattern} | Lowercase all chars |
| ${!prefix*} | Names matching prefix |
| ${!name[@]} | Array indices/keys |
| ${!var} | Indirect expansion |
| ${var@Q} | Quote for reuse |
| ${var@E} | Expand escape sequences |
| ${var@P} | Expand as prompt string |
| ${var@A} | Assignment statement form |
| ${var@a} | Attribute flags |
| ${var@U} | Uppercase all |
| ${var@u} | Uppercase first |
| ${var@L} | Lowercase all |
| ${var@K} | Key-value pairs (assoc arrays) |
Full details:
functions-parameters-expansions.md(section 2)
| Param | Meaning |
|---------------------|-------------------------------------------------|
| $0 | Script/shell name |
| $1..$9, ${10} | Positional parameters |
| $# | Number of positional parameters |
| $* | All positional params as single word (with IFS) |
| $@ | All positional params as separate words |
| "$*" | "$1c$2c..." where c = first char of IFS |
| "$@" | "$1" "$2" ... (preserves word boundaries) |
| $? | Exit status of last command |
| $$ | PID of the shell |
| $! | PID of last background command |
| $- | Current option flags |
| $_ | Last argument of previous command |
Full details:
shell-syntax-and-commands.md(section 3)
# If
if cmd; then ...; elif cmd; then ...; else ...; fi
# For
for var in words; do ...; done
for (( init; test; step )); do ...; done
# While / Until
while cmd; do ...; done
until cmd; do ...; done
# Case
case word in
pattern1|pattern2) commands ;; # break
pattern3) commands ;& # fall-through
pattern4) commands ;;& # test next
esac
# Select (menu)
select var in words; do ...; done
# Test
[[ expression ]] # Preferred (no word splitting/globbing)
(( expression )) # Arithmetic evaluation
# Grouping
{ commands; } # Current shell (note: space after {, ; before })
( commands ) # Subshell
[[ ]] and test/[ ])Full details:
bash-features.md(section 4)
| Operator | Test |
|-----------------------|----------------------------------|
| -e file | Exists |
| -f file | Regular file |
| -d file | Directory |
| -L file / -h file | Symlink |
| -s file | Non-zero size |
| -r file | Readable |
| -w file | Writable |
| -x file | Executable |
| -p file | Named pipe |
| -S file | Socket |
| -b file | Block device |
| -c file | Character device |
| -t fd | FD is terminal |
| -O file | Owned by effective UID |
| -G file | Owned by effective GID |
| -N file | Modified since last read |
| f1 -nt f2 | f1 newer than f2 |
| f1 -ot f2 | f1 older than f2 |
| f1 -ef f2 | Same inode |
| -v var | Variable is set |
| -R var | Variable is nameref |
| -z string | Zero length |
| -n string | Non-zero length |
| s1 == s2 | Equal (pattern match in [[ ]]) |
| s1 != s2 | Not equal |
| s1 < s2 | Less than (lexicographic) |
| s1 > s2 | Greater than (lexicographic) |
| s1 =~ regex | Regex match ([[ ]] only) |
| n1 -eq n2 | Numeric equal |
| n1 -ne n2 | Numeric not equal |
| n1 -lt n2 | Numeric less than |
| n1 -le n2 | Numeric less/equal |
| n1 -gt n2 | Numeric greater than |
| n1 -ge n2 | Numeric greater/equal |
Full details:
redirections-and-execution.md(section 1)
| Syntax | Operation |
|--------------------------------------|------------------------------------|
| cmd < file | Stdin from file |
| cmd > file | Stdout to file (truncate) |
| cmd >> file | Stdout to file (append) |
| cmd 2> file | Stderr to file |
| cmd &> file or cmd > file 2>&1 | Stdout+stderr to file |
| cmd &>> file or cmd >> file 2>&1 | Stdout+stderr append |
| cmd >| file | Force overwrite (noclobber) |
| cmd <<EOF | Here document |
| cmd <<-EOF | Here document (strip leading tabs) |
| cmd <<< "string" | Here string |
| cmd <&fd | Duplicate input FD |
| cmd >&fd | Duplicate output FD |
| cmd fd<&- | Close input FD |
| cmd fd>&- | Close output FD |
| cmd n<>file | Open for read+write on FD n |
| cmd {var}> file | Auto-assign FD to var |
Special filenames in redirections: /dev/fd/N, /dev/stdin, /dev/stdout, /dev/stderr, /dev/tcp/host/port, /dev/udp/host/port
Full details:
bash-features.md(section 6)
# Indexed arrays
declare -a arr=(one two three)
arr[0]="value"
arr+=(more items)
# Associative arrays (Bash 4.0+)
declare -A map=([key1]=val1 [key2]=val2)
map[key]="value"
# Access
${arr[0]} # Single element
${arr[@]} # All elements (separate words)
${arr[*]} # All elements (single word with IFS)
${#arr[@]} # Number of elements
${!arr[@]} # All indices/keys
${arr[@]:off:len} # Slice
# Unset
unset 'arr[2]' # Remove element (quote to prevent glob)
unset arr # Remove entire array
Steps 2-6 happen left-to-right simultaneously. Full details with word-count impacts: functions-parameters-expansions.md (section 3.5).
set OptionsFull details:
shell-builtins.md(section 2,setbuiltin)
| Option | Effect |
|------------------------|------------------------------------------------|
| set -e (errexit) | Exit on error (with exceptions) |
| set -u (nounset) | Error on unset variables |
| set -o pipefail | Pipeline fails if any command fails |
| set -x (xtrace) | Print commands before execution |
| set -f (noglob) | Disable filename expansion |
| set -n (noexec) | Read commands without executing (syntax check) |
| set -o posix | POSIX compliance mode |
| set -E (errtrace) | ERR trap inherited by functions |
| set -T (functrace) | DEBUG/RETURN traps inherited by functions |
shopt OptionsFull details:
shell-builtins.md(section 3,shoptoptions)
| Option | Effect |
|---------------------|-----------------------------------------------------------------|
| extglob | Extended patterns: ?(pat) *(pat) +(pat) @(pat) !(pat) |
| globstar | ** matches directories recursively |
| nullglob | Unmatched globs expand to nothing |
| failglob | Unmatched globs cause error |
| nocaseglob | Case-insensitive globbing |
| nocasematch | Case-insensitive case and [[ == ]] |
| dotglob | Globs match dotfiles |
| lastpipe | Last pipeline command runs in current shell |
| inherit_errexit | Command substitutions inherit errexit |
| assoc_expand_once | Expand associative array subscripts once |
Full details:
shell-builtins.md(trapbuiltin) andredirections-and-execution.md(section 5, Signals)
trap 'cleanup' EXIT # On shell exit
trap 'handle_err' ERR # On command error (with set -e)
trap 'on_debug' DEBUG # Before every command
trap 'on_return' RETURN # After function/sourced script returns
trap 'handle_int' INT # Ctrl-C
trap 'handle_term' TERM # kill signal
trap '' SIGNAL # Ignore signal
trap - SIGNAL # Reset to default
(( )) and $(( )))Full details:
bash-features.md(section 5)
All C-style operators: +, -, *, /, %, ** (exponent), <<, >>, &, |, ^, ~, !, &&, ||, <, >, <=, >=, ==, !=, =, +=, -=, *=, /=, %=, <<=, >>=, &=, |=, ^=, ++, --, expr?expr:expr (ternary), expr,expr (comma)
Bases: 0x (hex), 0 (octal), 0b (binary), base#number (arbitrary base 2-64)
Full details:
bash-features.md(section 8, Controlling the Prompt)
| Escape | Meaning |
|-----------|-------------------------------|
| \u | Username |
| \h | Hostname (short) |
| \H | Hostname (full) |
| \w | Working directory |
| \W | Basename of working directory |
| \d | Date (Day Mon Date) |
| \t | Time (HH:MM:SS 24hr) |
| \T | Time (HH:MM:SS 12hr) |
| \@ | Time (AM/PM) |
| \A | Time (HH:MM 24hr) |
| \D{fmt} | strftime format |
| \j | Number of jobs |
| \! | History number |
| \# | Command number |
| \$ | # if root, $ otherwise |
| \[ | Begin non-printing chars |
| \] | End non-printing chars |
| Term | Definition |
|----------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------|
| blank | Space or tab |
| word | Sequence of characters treated as a unit (no unquoted metacharacters) |
| token | A word or an operator |
| metacharacter | Unquoted: space, tab, newline, \|, &, ;, (, ), <, > |
| control operator | \|\|, &&, &, ;, ;;, ;&, ;;&, \|, \|&, (, ), newline |
| name/identifier | Letters, numbers, underscores; starts with letter or underscore |
| exit status | 0-255; 0 = success, 1 = general error, 2 = usage error, 126 = not executable, 127 = not found, 128+N = killed by signal N |
| special builtin | POSIX-designated builtins that have special properties (break, :, ., continue, eval, exec, exit, export, readonly, return, set, shift, trap, unset) |
#!/usr/bin/env bash
set -euo pipefail
tmpfile=$(mktemp) || exit 1
trap 'rm -f "$tmpfile"' EXIT
while IFS= read -r line; do
printf '%s\n' "$line"
done < "$file"
name="${1:-default}" # Use default if $1 unset/empty
name="${1:?'missing arg'}" # Exit with error if $1 unset/empty
# Lowercase / Uppercase
lower="${str,,}"
upper="${str^^}"
first_cap="${str^}"
# Trim prefix/suffix
filename="${path##*/}" # basename
dir="${path%/*}" # dirname
ext="${file##*.}" # extension
noext="${file%.*}" # remove extension
for item in "${arr[@]}"; do echo "$item"; done # values
for i in "${!arr[@]}"; do echo "$i: ${arr[$i]}"; done # index: value
diff <(sort file1) <(sort file2)
while IFS= read -r line; do ...; done < <(cmd) # avoid subshell
declare -A map
if [[ -v map["key"] ]]; then echo "exists"; fi
# BAD: word splits and globs if file contains spaces or wildcards
for f in $files; do rm $f; done
# GOOD: always quote variable expansions
for f in "${files[@]}"; do rm "$f"; done
[ ] vs [[ ]]# BAD: word splitting inside [ ] can break with spaces in $var
[ $var == "hello" ] # also: == is not POSIX in [ ]
# GOOD: [[ ]] prevents word splitting and supports == and =~
[[ $var == "hello" ]]
$@ vs $* quoting# BAD: loses word boundaries
for arg in $@; do echo "$arg"; done
# GOOD: preserves each argument as a separate word
for arg in "$@"; do echo "$arg"; done
# "$*" joins all args into ONE word (separated by first char of IFS)
set -e doesn't trigger everywhereset -e
# These do NOT cause exit on failure:
if false; then :; fi # test in 'if'
false || true # LHS of ||
false && true # LHS of &&
false | true # pipeline (without pipefail)
! false # negated command
; before } in brace groups# BAD: syntax error
{ echo "hello" }
# GOOD: semicolon (or newline) required before }
{ echo "hello"; }
arr=(one two three)
# BAD: expands $arr (element 0) then appends literal [1]
echo $arr[1] # prints: one[1]
# GOOD: braces required for array subscript
echo "${arr[1]}" # prints: two
declare -A for associative arrays# BAD: creates an indexed array, keys treated as arithmetic (0)
map=([foo]=1 [bar]=2)
echo "${map[foo]}" # prints: 2 (both keys evaluated to index 0)
# GOOD: must declare associative arrays explicitly
declare -A map=([foo]=1 [bar]=2)
echo "${map[foo]}" # prints: 1
= vs == in the wrong context# In [ ] / test: use = (POSIX). == works in Bash but is not portable.
[ "$a" = "$b" ]
# In [[ ]]: both = and == work; RHS is a pattern (quote for literal match)
[[ "$a" == "$b" ]] # literal match (RHS quoted)
[[ "$a" == "$b"* ]] # glob pattern (unquoted * appended)
[[ "$a" == $b ]] # if $b contains *, it's a pattern (RHS unquoted)
# BAD: read runs in subshell, $var is lost after pipeline
echo "hello" | read var
echo "$var" # empty
# GOOD: use process substitution or lastpipe
read var < <(echo "hello")
echo "$var" # hello
# OR: enable lastpipe (non-interactive, no job control)
shopt -s lastpipe
echo "hello" | read var
local variables use dynamic scopinginner() { echo "$x"; }
outer() { local x=42; inner; }
outer # prints: 42 (inner sees outer's local!)
# This is dynamic scoping, not lexical. Any function called
# from a scope with a local variable sees that variable.
read -r var <<< "hello"
printf '%s' "$var" | xxd # contains "hello\n" — trailing newline added
# Use printf instead when exact bytes matter:
read -r var < <(printf '%s' "hello")
declare -i silently evaluates strings as arithmeticdeclare -i num
num="1+1"
echo "$num" # prints: 2 (string was evaluated!)
# WARNING: with untrusted input this is a code injection vector:
# declare -i x; x="a]$(cmd)" would execute cmd
trap EXIT is reset in subshellstrap 'echo cleanup' EXIT
(echo "subshell") # EXIT trap does NOT fire here
# Subshells inherit the parent's traps but reset EXIT/ERR/DEBUG/RETURN.
# If you need cleanup in a subshell, set a new trap inside it.
development
Textual TUI framework documentation reference
testing
Use when creating new skills, editing existing skills, structuring SKILL.md files, writing skill frontmatter, testing skills with subagents, deploying skills, or verifying skills work before deployment
development
Use when building transparent remote procedure calls, distributed computing, or remote object proxying in Python with RPyC. Use when asked about rpyc.connect, rpyc.Service, netref proxies, async_(), BgServingThread, SSLAuthenticator, DeployedServer, or rpyc_classic.py.
tools
Use when choosing Python libraries for a task, when writing new Python code that needs dependencies, when reviewing Python imports for non-preferred libraries, or when unsure which library to use for JSON, HTTP, logging, TOML, YAML, compression, database access, testing, or CLI tools