chore: tune challenge matrix and debug prompts

This commit is contained in:
lda
2026-06-29 19:25:35 +07:00 Verified
parent 5f520d74bd
commit e0315f295c
10 changed files with 27 additions and 18 deletions
@@ -127,6 +127,7 @@ def test_challenge_prompt_is_identical_across_profiles(tmp_path: Path) -> None:
assert ".agent/skills" in rendered[InstructionProfile.SKILLS].text
assert "inspect broader repository" in rendered[InstructionProfile.ALL].text
assert "genuinely blocked" in rendered[InstructionProfile.DEBUG].text
assert "debug profile only" in rendered[InstructionProfile.DEBUG].text
assert "ux_issues_found" in rendered[InstructionProfile.DEBUG].text
@@ -950,9 +950,8 @@ def test_browser_click_wrapper_produces_expected_paths_and_command_prefix() -> N
)
assert BROWSER_CLICK_DEF.default_prompt.name == "challenge-prompt.md"
assert BROWSER_CLICK_DEF.default_prompt.parent.name == "browser_click_challenge"
assert "ux_issues_found: []" in BROWSER_CLICK_DEF.default_prompt.read_text(
encoding="utf-8"
)
prompt = BROWSER_CLICK_DEF.default_prompt.read_text(encoding="utf-8")
assert "Debug profile only: include ux_issues_found here" in prompt
def test_generic_runner_can_be_configured_with_fake_challenge_and_fake_opencode(
@@ -40,7 +40,7 @@ def test_report_challenge_prompt_requires_full_product_lifecycle() -> None:
assert "render_markdown_report" in prompt
assert "deployment" in prompt.lower()
assert "run_id" in prompt
assert "ux_issues_found: []" in prompt
assert "Debug profile only: include ux_issues_found here" in prompt
def test_report_challenge_workspace_template_contains_safe_input_files(