test: smoke tests for validate_paths and print_report

[bipinhcs11]

What and why

validate_paths and print_report are the two functions wired to the skill-gen validate subcommand — the entry point developers use to audit SKILL.md files before committing. Both were exported from validate.py but had zero test coverage. The existing test_validate.py covered validate() and validate_file() only; the CLI-facing layer was untested.

What's covered (14 new tests)

TestValidatePaths (9 tests)

Test	What it pins
test_explicit_valid_file_returns_single_ok_result	Single-file happy path
test_explicit_invalid_file_returns_single_error_result	Single-file error path
test_nonexistent_file_returns_error_result	Missing file → error result, not a crash
test_empty_path_list_returns_empty_list	Empty input → empty list
test_directory_finds_all_skill_mds_recursively	Directory glob finds 3 deeply-nested SKILL.mds
test_directory_with_no_skill_mds_returns_empty	Directory with no SKILL.md → empty list
test_mix_of_file_and_directory_combined	Explicit file + directory counted together
test_results_have_path_attribute	Each result carries a path field
test_three_reference_skill_dirs_all_ok	Regression: the three artifact-3 reference skills in this repo must continue to pass validate_paths cleanly; a future edit that breaks them will fail this test

TestPrintReport (5 tests)

Test	What it pins
test_all_ok_returns_zero	Exit code 0 when no errors
test_any_error_returns_one	Exit code 1 when any error present
test_output_contains_summary_line	Summary line with "validated" / "ok" always printed
test_empty_results_returns_zero	Empty input → exit 0, no crash
test_error_count_in_summary	"error" appears in output when errors exist

Note on skills/skill-generator/SKILL.md

The skills/ tree contains a fourth SKILL.md — the Claude Agent skill spec at skills/skill-generator/SKILL.md. This file uses a different frontmatter schema (name:/description: rather than the artifact-3 fields) so it legitimately fails validate. The regression test therefore targets the three artifact-3 reference skill directories explicitly rather than the whole skills/ tree.

Suite size

172 → 186 tests, all green.

python3 -m unittest tests/test_validate.py -v  # 51 passed, 0 failed
python3 -m unittest discover tests/           # 186 passed, 0 failed

Follow-ups (not in this PR)

• cli.py has no tests. build_parser() is a pure function; smoke tests for all 11 subcommand argument shapes would catch future wiring bugs.
• PRs test: smoke tests for update.py — last uncovered module #12 / test: smoke tests for update.py (Phase 2 incremental updater) #14 / test: smoke tests for update.py (was the only uncovered module) #17 / test: smoke tests for update.py pure functions (Phase 2 updater) #19 all add tests/test_update.py; one of them should be merged and the rest closed.

---

Generated by Claude Code