test: add --help scenario tests for cut, echo, head, printf, sed, sort, tail, wc

[thieman]

Problem

T-1: Missing --help scenario tests for 8 commands

RULES.md requires: "Every supported flag MUST have correctness tests."

These 8 commands had no scenario tests verifying that --help (and -h where supported) exits with the correct code and prints usage to stdout:

• cut — --help only
• echo — no --help flag; documents that it prints the literal string
• head — --help and -h
• printf — --help only (exits 2 per implementation)
• sed — --help and -h
• sort — --help and -h
• tail — --help and -h
• wc — --help only

Fix

Added help/help_flag.yaml (and help/short_help_flag.yaml for -h variants) for each command. Tests verify:

• Correct exit code (0 for all except printf which exits 2)
• Usage information printed to stdout (via stdout_contains)
• No stderr output

All tests are marked skip_assert_against_bash: true since bash handles these differently.

Test plan

• go test ./tests/... -run TestShellScenarios passes with all 12 new scenario files

🤖 Generated with Claude Code

[thieman]

@codex review this PR

[thieman]

Self-Review: PR #151 — Add help flag scenario tests

Scope reviewed: 12 new YAML scenario test files across 8 commands (cut, echo, head, printf, sed, sort, tail, wc).

Overall assessment: needs minor fixes — the tests correctly capture help-flag behavior and all pass. Two findings below.

Coverage Summary

Code path	Scenario test	Go test	Status
cut --help → stdout, exit 0	cut/help/help_flag.yaml	—	Covered
echo --help → prints literal text, exit 0	echo/help/help_flag.yaml	—	Covered
head --help → stdout, exit 0	head/help/help_flag.yaml	—	Covered
head -h → stdout, exit 0	head/help/short_help_flag.yaml	—	Covered
printf --help → stdout, exit 2	printf/help/help_flag.yaml	—	Covered
sed --help → stdout, exit 0	sed/help/help_flag.yaml	—	Covered
sed -h → stdout, exit 0	sed/help/short_help_flag.yaml	—	Covered
sort --help → stdout, exit 0	sort/help/help_flag.yaml	—	Covered
sort -h → stdout, exit 0	sort/help/short_help_flag.yaml	—	Covered
tail --help → stdout, exit 0	tail/help/help_flag.yaml	—	Covered
tail -h → stdout, exit 0	tail/help/short_help_flag.yaml	—	Covered
wc --help → stdout, exit 0	wc/help/help_flag.yaml	—	Covered

Findings

#	Priority	File	Finding
1		tests/scenarios/cmd/echo/help/help_flag.yaml:3	skip_assert_against_bash is set but bash behavior matches rshell
2		tests/scenarios/cmd/printf/help/help_flag.yaml:9	stdout_contains assertion is too loose

Positive Observations

• All 12 scenarios correctly align with the actual builtin implementations (verified by reading source and running tests).
• exit_code: 2 for printf --help correctly matches the implementation.
• Commands without a -h short flag (cut, wc) correctly omit short_help_flag.yaml.
• All files follow the project YAML conventions (|+ block scalars, expect.stderr exact match).

[thieman]

Iteration 1 self-review result: 2 findings — 1× P2, 1× P3. No security or correctness issues. The P2 is an unnecessary skip_assert_against_bash: true in the echo scenario (bash behavior matches); the P3 is a loose stdout_contains assertion in the printf scenario.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: bfee7e45e4

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Re-enable bash comparison for this echo --help scenario

This test is explicitly asserting bash-equivalent behavior (echo --help is treated as a literal string), but skip_assert_against_bash: true prevents it from running in TestShellScenariosAgainstBash. I verified in bash that echo --help exits 0, writes --help to stdout, and leaves stderr empty, so the current expectations are compatible; skipping here removes useful compatibility coverage and can hide regressions.

Useful? React with 👍 / 👎.