set -euo pipefail as a Script Header

#!/bin/bash
set -euo pipefail

name="demo script"
echo "running $name safely"

#!/bin/bash
set -o pipefail

if false | true; then
    echo "unreachable with pipefail active"
else
    echo "pipefail correctly noticed the earlier failure"
fi

#!/bin/bash
(
    set -u
    echo "testing just -u in isolation"
)
(
    set -e
    echo "testing just -e in isolation"
)
echo "both isolated tests completed without crashing the main script"

#!/bin/bash
set -euo pipefail

process_item() {
    local item=$1
    echo "processing: $item"
}

for item in alpha beta gamma; do
    process_item "$item"
done
echo "all items processed successfully"

1. Adding set -euo pipefail to an existing large script and being surprised when it suddenly "breaks", when really it's exposing bugs (unset variables, ignored failures) that were already there.
2. Forgetting that set -e's exemptions (inside if/while conditions, non-last pipeline stages without pipefail) still apply even with the combined flags.
3. Not testing scripts under set -euo pipefail incrementally, making it hard to tell which of the three flags caused a given failure.

• set -euo pipefail combines exit-on-error (-e), error-on-unset-variable (-u), and pipeline-failure-propagation (-o pipefail) into one line.
• It's a widely recommended defensive header for non-trivial bash scripts, though it doesn't catch every possible bug.
• Adding it to a pre-existing script often surfaces previously silent bugs rather than introducing new ones.
• It's still worth explicit error handling for the specific failures you most care about, rather than relying on strict mode alone.