claude-dev-env 1.59.0 → 1.61.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (81) hide show
  1. package/CLAUDE.md +4 -0
  2. package/audit-rubrics/category_rubrics/category-b-selector-engine-compat.md +1 -1
  3. package/audit-rubrics/category_rubrics/category-e-dead-code.md +1 -0
  4. package/audit-rubrics/category_rubrics/category-f-silent-failures.md +1 -1
  5. package/audit-rubrics/category_rubrics/category-o-docstring-vs-impl-drift.md +1 -1
  6. package/audit-rubrics/prompts/category-b-selector-engine-compat.md +2 -2
  7. package/audit-rubrics/prompts/category-e-dead-code.md +17 -4
  8. package/audit-rubrics/prompts/category-f-silent-failures.md +1 -0
  9. package/docs/CODE_RULES.md +2 -2
  10. package/hooks/blocking/code_rules_annotations_length.py +189 -10
  11. package/hooks/blocking/code_rules_dead_module_constant.py +321 -0
  12. package/hooks/blocking/code_rules_duplicate_body.py +152 -0
  13. package/hooks/blocking/code_rules_enforcer.py +38 -15
  14. package/hooks/blocking/code_rules_orphan_css_class.py +196 -0
  15. package/hooks/blocking/code_rules_typeddict_stub.py +172 -0
  16. package/hooks/blocking/config/__init__.py +5 -0
  17. package/hooks/blocking/config/verified_commit_constants.py +118 -0
  18. package/hooks/blocking/destructive_command_blocker.py +483 -61
  19. package/hooks/blocking/test_code_rules_enforcer_annotations.py +240 -0
  20. package/hooks/blocking/test_code_rules_enforcer_cap_meta.py +1 -0
  21. package/hooks/blocking/test_code_rules_enforcer_cross_skill_duplicate.py +146 -0
  22. package/hooks/blocking/test_code_rules_enforcer_dead_module_constant.py +188 -0
  23. package/hooks/blocking/test_code_rules_enforcer_dispatch_wiring.py +82 -0
  24. package/hooks/blocking/test_code_rules_enforcer_orphan_css_class.py +196 -0
  25. package/hooks/blocking/test_code_rules_enforcer_zero_payload_alias.py +415 -0
  26. package/hooks/blocking/test_code_rules_enforcer_zero_payload_alias_hook_routing.py +156 -0
  27. package/hooks/blocking/test_destructive_command_blocker.py +213 -0
  28. package/hooks/blocking/test_verdict_directory_write_blocker.py +720 -0
  29. package/hooks/blocking/test_verification_verdict_store.py +490 -0
  30. package/hooks/blocking/test_verified_commit_gate.py +495 -0
  31. package/hooks/blocking/test_verified_commit_message_accuracy_blocker.py +131 -0
  32. package/hooks/blocking/test_verifier_verdict_minter.py +193 -0
  33. package/hooks/blocking/verdict_directory_write_blocker.py +667 -0
  34. package/hooks/blocking/verification_verdict_store.py +686 -0
  35. package/hooks/blocking/verified_commit_gate.py +535 -0
  36. package/hooks/blocking/verified_commit_message_accuracy_blocker.py +152 -0
  37. package/hooks/blocking/verifier_verdict_minter.py +221 -0
  38. package/hooks/diagnostic/test_hook_log_extractor.py +3 -3
  39. package/hooks/hooks.json +43 -1
  40. package/hooks/hooks_constants/blocking_check_limits.py +1 -0
  41. package/hooks/hooks_constants/code_rules_enforcer_constants.py +6 -0
  42. package/hooks/hooks_constants/dead_module_constant_constants.py +20 -0
  43. package/hooks/hooks_constants/destructive_command_segment_constants.py +15 -0
  44. package/hooks/hooks_constants/duplicate_function_body_constants.py +22 -5
  45. package/hooks/hooks_constants/orphan_css_class_constants.py +40 -0
  46. package/hooks/hooks_constants/precommit_code_rules_gate_constants.py +1 -1
  47. package/hooks/validation/mypy_validator.py +59 -7
  48. package/hooks/validation/test_mypy_validator.py +94 -0
  49. package/package.json +1 -1
  50. package/rules/file-global-constants.md +7 -1
  51. package/rules/no-cross-skill-duplicate-helpers.md +29 -0
  52. package/rules/orphan-css-class.md +23 -0
  53. package/skills/_shared/pr-loop/scripts/preflight_worktree.py +392 -0
  54. package/skills/_shared/pr-loop/scripts/skills_pr_loop_constants/preflight_constants.py +70 -0
  55. package/skills/_shared/pr-loop/scripts/test_preflight_worktree.py +263 -0
  56. package/skills/autoconverge/SKILL.md +54 -17
  57. package/skills/autoconverge/reference/closing-report.md +59 -17
  58. package/skills/autoconverge/workflow/aggregate_runs.py +371 -0
  59. package/skills/autoconverge/workflow/autoconverge_report_constants/render_report_constants.py +192 -76
  60. package/skills/autoconverge/workflow/converge.clean-audit.test.mjs +76 -0
  61. package/skills/autoconverge/workflow/converge.contract.test.mjs +395 -206
  62. package/skills/autoconverge/workflow/converge.mjs +520 -57
  63. package/skills/autoconverge/workflow/convergence_summary.py +110 -0
  64. package/skills/autoconverge/workflow/fixtures/wf_run/subagents/workflows/wf_881252e6-700/agent-ab1c2d3e4f5a6b7c8.jsonl +2 -0
  65. package/skills/autoconverge/workflow/fixtures/wf_run/workflows/wf_881252e6-700.json +7 -0
  66. package/skills/autoconverge/workflow/render_report.py +488 -397
  67. package/skills/autoconverge/workflow/test_aggregate_runs.py +134 -0
  68. package/skills/autoconverge/workflow/test_convergence_summary.py +132 -0
  69. package/skills/autoconverge/workflow/test_render_report.py +518 -259
  70. package/skills/pr-converge/reference/per-tick.md +28 -8
  71. package/skills/rebase/SKILL.md +2 -4
  72. package/system-prompts/software-engineer.xml +2 -6
  73. package/hooks/blocking/content_search_to_zoekt_redirector.py +0 -59
  74. package/hooks/blocking/content_search_zoekt_bash_block_reason.py +0 -25
  75. package/hooks/blocking/content_search_zoekt_block_payload.py +0 -21
  76. package/hooks/blocking/content_search_zoekt_indexed_paths.py +0 -24
  77. package/hooks/blocking/content_search_zoekt_indexed_roots_config.py +0 -131
  78. package/hooks/blocking/content_search_zoekt_redirect_guidance.py +0 -52
  79. package/hooks/blocking/test_content_search_to_zoekt_redirector_integration.py +0 -61
  80. package/hooks/blocking/test_content_search_to_zoekt_redirector_unit.py +0 -92
  81. package/hooks/blocking/test_content_search_zoekt_indexed_roots_config.py +0 -102
@@ -15,59 +15,20 @@ FIXTURE_DIR = Path(__file__).resolve().parent / "fixtures" / "wf_run"
15
15
  FIXTURE_JOURNAL = FIXTURE_DIR / "workflows" / "wf_881252e6-700.json"
16
16
 
17
17
  EXPECTED_TOTAL_FINDINGS = 15
18
- EXPECTED_CRITICAL_COUNT = 0
19
- EXPECTED_MINOR_COUNT = 15
20
18
  EXPECTED_FIX_COMMIT_COUNT = 2
21
19
  EXPECTED_GENERATED_DATE = "2026-06-13"
22
- EXPECTED_FINDINGS_BY_ROUND = {1: 11, 2: 2, 3: 2, 4: 0}
23
- EXPECTED_FINDINGS_BY_THEME = {"src/exports": 11, "src/logging": 2, "src/web": 2}
20
+ EXPECTED_ROUND_COUNT = 4
24
21
 
25
22
 
26
- def test_load_run_data_aggregate_counts() -> None:
27
- """Should parse the fixture journal and transcripts into correct aggregate counts."""
28
- run_data = render_report.load_run_data(FIXTURE_JOURNAL, Path("."))
29
-
30
- assert run_data.total_finding_count == EXPECTED_TOTAL_FINDINGS
31
- assert run_data.critical_finding_count == EXPECTED_CRITICAL_COUNT
32
- assert run_data.minor_finding_count == EXPECTED_MINOR_COUNT
33
- assert run_data.fix_commit_count == EXPECTED_FIX_COMMIT_COUNT
34
- assert run_data.generated_date == EXPECTED_GENERATED_DATE
35
-
36
-
37
- def test_load_run_data_by_round_counts() -> None:
38
- """Should assign findings to rounds by workflowProgress position boundary."""
39
- run_data = render_report.load_run_data(FIXTURE_JOURNAL, Path("."))
40
-
41
- for each_round, expected_count in EXPECTED_FINDINGS_BY_ROUND.items():
42
- actual_count = run_data.finding_count_by_round.get(each_round, 0)
43
- assert actual_count == expected_count, (
44
- f"Round {each_round}: expected {expected_count}, got {actual_count}"
45
- )
46
-
47
-
48
- def test_load_run_data_by_theme_counts() -> None:
49
- """Should group distinct findings by the first two path segments."""
50
- run_data = render_report.load_run_data(FIXTURE_JOURNAL, Path("."))
51
-
52
- assert len(run_data.finding_count_by_theme) == len(EXPECTED_FINDINGS_BY_THEME)
53
- for each_theme, expected_count in EXPECTED_FINDINGS_BY_THEME.items():
54
- actual_count = run_data.finding_count_by_theme.get(each_theme, 0)
55
- assert actual_count == expected_count, (
56
- f"Theme {each_theme}: expected {expected_count}, got {actual_count}"
57
- )
58
-
59
-
60
- def test_cli_end_to_end(tmp_path: Path) -> None:
61
- """Should exit 0, print the output path, and write HTML with expected substrings."""
62
- out_path = tmp_path / "report.html"
23
+ def _render_cli(journal_path: Path, out_path: Path) -> subprocess.CompletedProcess[str]:
24
+ """Run the render_report CLI against a journal and return the completed process."""
63
25
  render_script = Path(__file__).resolve().parent / "render_report.py"
64
-
65
- completed = subprocess.run(
26
+ return subprocess.run(
66
27
  [
67
28
  sys.executable,
68
29
  str(render_script),
69
30
  "--journal",
70
- str(FIXTURE_JOURNAL),
31
+ str(journal_path),
71
32
  "--out",
72
33
  str(out_path),
73
34
  "--pr",
@@ -76,58 +37,327 @@ def test_cli_end_to_end(tmp_path: Path) -> None:
76
37
  "7c2f420c4d5b7c83aa47f93d99a0f1420e3373c4",
77
38
  "--rounds",
78
39
  "4",
79
- "--repo",
80
- ".",
81
40
  ],
82
41
  capture_output=True,
83
42
  text=True,
84
43
  )
85
44
 
86
- assert completed.returncode == 0, f"CLI failed:\n{completed.stderr}"
87
45
 
88
- printed_path = completed.stdout.strip()
89
- assert printed_path == str(out_path), (
90
- f"Expected stdout {out_path!r}, got {printed_path!r}"
46
+ def _copy_run_tree_without_summary_entry(destination_root: Path) -> Path:
47
+ """Copy the fixture run tree, dropping the convergence-summary workflowProgress entry.
48
+
49
+ Returns the path to the copied journal whose summarizer entry has been removed.
50
+ """
51
+ shutil.copytree(FIXTURE_DIR, destination_root)
52
+ journal_destination = destination_root / "workflows" / FIXTURE_JOURNAL.name
53
+ journal = json.loads(journal_destination.read_text(encoding="utf-8"))
54
+ journal["workflowProgress"] = [
55
+ each_entry
56
+ for each_entry in journal["workflowProgress"]
57
+ if each_entry.get("label") != render_report.LABEL_CONVERGENCE_SUMMARY
58
+ ]
59
+ journal_destination.write_text(json.dumps(journal, indent=2), encoding="utf-8")
60
+ return journal_destination
61
+
62
+
63
+ def test_load_run_data_aggregate_counts() -> None:
64
+ """Should parse the fixture journal and transcripts into correct aggregate counts."""
65
+ run_data = render_report.load_run_data(FIXTURE_JOURNAL)
66
+
67
+ assert run_data.total_finding_count == EXPECTED_TOTAL_FINDINGS
68
+ assert run_data.fix_commit_count == EXPECTED_FIX_COMMIT_COUNT
69
+ assert run_data.generated_date == EXPECTED_GENERATED_DATE
70
+ assert len(run_data.all_distinct_findings) == EXPECTED_TOTAL_FINDINGS
71
+
72
+
73
+ def test_load_run_data_parses_convergence_summary() -> None:
74
+ """Should locate the convergence-summary entry and parse its StructuredOutput."""
75
+ run_data = render_report.load_run_data(FIXTURE_JOURNAL)
76
+
77
+ assert run_data.convergence_summary is not None
78
+ verdict_line = run_data.convergence_summary["verdictLine"]
79
+ issue_classes = run_data.convergence_summary["issueClasses"]
80
+ assert isinstance(verdict_line, str) and verdict_line
81
+ assert isinstance(issue_classes, list) and len(issue_classes) == 3
82
+
83
+
84
+ def test_load_run_data_carries_category_on_findings() -> None:
85
+ """Should default each finding's category to 'bug' when the raw dict omits it."""
86
+ run_data = render_report.load_run_data(FIXTURE_JOURNAL)
87
+
88
+ assert all(
89
+ each_finding.category == render_report.CATEGORY_BUG
90
+ for each_finding in run_data.all_distinct_findings
91
91
  )
92
92
 
93
+
94
+ def test_cli_renders_verdict_banner_with_python_computed_vsub(tmp_path: Path) -> None:
95
+ """Should render the verdict banner with verdictLine and a Python-computed vsub."""
96
+ out_path = tmp_path / "report.html"
97
+
98
+ completed = _render_cli(FIXTURE_JOURNAL, out_path)
99
+
100
+ assert completed.returncode == 0, f"CLI failed:\n{completed.stderr}"
101
+ assert completed.stdout.strip() == str(out_path)
93
102
  assert out_path.exists(), "Output HTML file was not written"
103
+
94
104
  html_content = out_path.read_text(encoding="utf-8")
105
+ assert "PR #211 Convergence Summary" in html_content
106
+ assert 'class="verdict"' in html_content
107
+ assert 'class="vtext"' in html_content
108
+ assert "Converged in 4 rounds; 3 distinct issue classes were caught and fixed." in (
109
+ html_content
110
+ )
111
+ assert 'class="vsub"' in html_content
112
+ assert "2 fix commits" in html_content
113
+ assert "final commit 7c2f420c" in html_content
95
114
 
96
- expected_substrings = [
97
- "PR #211 Convergence Insights",
98
- "at-a-glance",
99
- "Findings by severity",
100
- "Findings by round",
101
- "Tests added per round",
102
- "Findings by theme",
103
- "Banned identifier",
104
- "result",
105
- "in test",
106
- "Converged",
107
- "7c2f420c",
108
- ]
109
- for each_substring in expected_substrings:
110
- assert each_substring in html_content, (
111
- f"Expected substring not found in HTML: {each_substring!r}"
112
- )
113
115
 
114
- minor_card_count = html_content.count('class="bug-card minor"')
115
- assert minor_card_count == EXPECTED_MINOR_COUNT, (
116
- f"Expected {EXPECTED_MINOR_COUNT} minor cards, found {minor_card_count}"
117
- )
116
+ def test_cli_renders_problem_and_fix_scene_cards(tmp_path: Path) -> None:
117
+ """Should draw problem and fix scene cards with trigger, result, and caption."""
118
+ out_path = tmp_path / "report-scenes.html"
118
119
 
120
+ completed = _render_cli(FIXTURE_JOURNAL, out_path)
121
+ assert completed.returncode == 0, f"CLI failed:\n{completed.stderr}"
119
122
 
120
- def test_html_contains_no_hedging_words(tmp_path: Path) -> None:
121
- """Should produce HTML with no hedging language anywhere in the rendered text."""
122
- out_path = tmp_path / "report-hedge.html"
123
- render_script = Path(__file__).resolve().parent / "render_report.py"
123
+ html_content = out_path.read_text(encoding="utf-8")
124
+ assert 'class="pf-grid"' in html_content
125
+ assert 'class="pf problem"' in html_content
126
+ assert 'class="pf fix"' in html_content
127
+ assert "export stops at batch 90 of 100" in html_content
128
+ assert "starts again at batch 1" in html_content
129
+ assert "continues at batch 91" in html_content
130
+ assert 'class="res-bad"' in html_content
131
+ assert 'class="res-good"' in html_content
132
+ assert "began again" in html_content
133
+
134
+
135
+ def test_render_issue_class_panels_for_each_medium() -> None:
136
+ """Should draw before/after panels: a code panel and a terminal panel per medium."""
137
+ convergence_summary = {
138
+ "verdictLine": "Converged.",
139
+ "problemScenes": [],
140
+ "fixScenes": [],
141
+ "issueClasses": [
142
+ {
143
+ "plainName": "A missing return type",
144
+ "count": 3,
145
+ "severity": "P2",
146
+ "category": "code-standard",
147
+ "status": "fixed",
148
+ "cause": "Tests did not declare their return type.",
149
+ "medium": "code",
150
+ "beforeLines": ["def test_x():"],
151
+ "afterLines": ["def test_x() -> None:"],
152
+ },
153
+ {
154
+ "plainName": "An install that did nothing",
155
+ "count": 1,
156
+ "severity": "P1",
157
+ "category": "bug",
158
+ "status": "fixed",
159
+ "cause": "The command skipped the install.",
160
+ "medium": "terminal",
161
+ "beforeLines": ["~ $ install", "(no output)"],
162
+ "afterLines": ["~ $ install", "Installed."],
163
+ },
164
+ ],
165
+ }
166
+
167
+ panels_html = render_report._render_issue_class_panels(convergence_summary)
168
+
169
+ assert 'class="code-panel"' in panels_html
170
+ assert "def test_x() -> None:" in panels_html
171
+ assert 'class="terminal"' in panels_html
172
+ assert 'class="term-bar"' in panels_html
173
+ assert "Installed." in panels_html
174
+ assert 'class="term-grid"' in panels_html
175
+ assert 'class="bug-head"' in panels_html
176
+ assert "A missing return type" in panels_html
177
+ assert "An install that did nothing" in panels_html
178
+ assert "3 findings" in panels_html
179
+ assert "1 finding" in panels_html
180
+
181
+
182
+ def test_render_issue_class_panels_draws_text_panel_for_text_medium() -> None:
183
+ """Should draw a text panel holding the supplied lines when the medium is text."""
184
+ convergence_summary = {
185
+ "verdictLine": "Converged.",
186
+ "problemScenes": [],
187
+ "fixScenes": [],
188
+ "issueClasses": [
189
+ {
190
+ "plainName": "A plain-text symptom",
191
+ "count": 1,
192
+ "severity": "P2",
193
+ "category": "bug",
194
+ "status": "fixed",
195
+ "cause": "A grounded cause sentence.",
196
+ "medium": "text",
197
+ "beforeLines": ["pages reloaded every visit"],
198
+ "afterLines": ["pages reuse the saved copy"],
199
+ }
200
+ ],
201
+ }
202
+
203
+ panels_html = render_report._render_issue_class_panels(convergence_summary)
124
204
 
125
- subprocess.run(
205
+ assert 'class="text-panel"' in panels_html
206
+ assert "pages reloaded every visit" in panels_html
207
+ assert "pages reuse the saved copy" in panels_html
208
+ assert 'class="terminal"' not in panels_html
209
+ assert 'class="code-panel"' not in panels_html
210
+
211
+
212
+ def test_cli_renders_cause_line_with_severity_parenthetical(tmp_path: Path) -> None:
213
+ """Should render a cause line carrying the plain cause and a muted parenthetical."""
214
+ out_path = tmp_path / "report-cause.html"
215
+
216
+ completed = _render_cli(FIXTURE_JOURNAL, out_path)
217
+ assert completed.returncode == 0, f"CLI failed:\n{completed.stderr}"
218
+
219
+ html_content = out_path.read_text(encoding="utf-8")
220
+ assert 'class="cause"' in html_content
221
+ assert "which the project's type checker wants" in html_content
222
+ assert "P2" in html_content
223
+ assert "code standard" in html_content
224
+ assert "×7" in html_content
225
+ assert "fixed" in html_content
226
+
227
+
228
+ def test_render_issue_class_panels_omitted_when_lines_empty() -> None:
229
+ """Should draw only the cause line when both before and after lines are empty."""
230
+ convergence_summary = {
231
+ "verdictLine": "Converged.",
232
+ "problemScenes": [],
233
+ "fixScenes": [],
234
+ "issueClasses": [
235
+ {
236
+ "plainName": "A cause-only class",
237
+ "count": 1,
238
+ "severity": "P2",
239
+ "category": "code-standard",
240
+ "status": "fixed",
241
+ "cause": "Nothing visual to show.",
242
+ "medium": "text",
243
+ "beforeLines": [],
244
+ "afterLines": [],
245
+ }
246
+ ],
247
+ }
248
+
249
+ panels_html = render_report._render_issue_class_panels(convergence_summary)
250
+
251
+ assert 'class="term-grid"' not in panels_html
252
+ assert 'class="bug-head"' in panels_html
253
+ assert "A cause-only class" in panels_html
254
+ assert 'class="cause"' in panels_html
255
+ assert "Nothing visual to show." in panels_html
256
+
257
+
258
+ def test_render_issue_class_panels_clean_state_when_no_classes() -> None:
259
+ """Should render a clean-state line, not an empty section, when no classes exist."""
260
+ convergence_summary = {
261
+ "verdictLine": "Converged with no issues caught.",
262
+ "problemScenes": [],
263
+ "fixScenes": [],
264
+ "issueClasses": [],
265
+ }
266
+
267
+ panels_html = render_report._render_issue_class_panels(convergence_summary)
268
+
269
+ assert 'class="term-grid"' not in panels_html
270
+ assert "No issues were caught" in panels_html
271
+
272
+
273
+ def test_cli_merges_run_stats_lead_into_caught_section(tmp_path: Path) -> None:
274
+ """Should lead the caught section with run stats and omit any timeline section."""
275
+ out_path = tmp_path / "report-caught-lead.html"
276
+
277
+ completed = _render_cli(FIXTURE_JOURNAL, out_path)
278
+ assert completed.returncode == 0, f"CLI failed:\n{completed.stderr}"
279
+
280
+ html_content = out_path.read_text(encoding="utf-8")
281
+ assert "What was caught" in html_content
282
+ assert "3 bug classes" in html_content
283
+ assert "15 findings in all" in html_content
284
+ assert "caught and fixed across 4 rounds" in html_content
285
+ assert "2 fix commits" in html_content
286
+ assert "How it converged" not in html_content
287
+ assert 'class="timeline"' not in html_content
288
+ assert 'class="tstep' not in html_content
289
+
290
+
291
+ def test_cli_includes_collapsed_appendix(tmp_path: Path) -> None:
292
+ """Should include a collapsed details appendix listing every distinct finding."""
293
+ out_path = tmp_path / "report-appendix.html"
294
+
295
+ completed = _render_cli(FIXTURE_JOURNAL, out_path)
296
+ assert completed.returncode == 0, f"CLI failed:\n{completed.stderr}"
297
+
298
+ html_content = out_path.read_text(encoding="utf-8")
299
+ assert '<details class="appendix"' in html_content
300
+ assert f"Raw findings ({EXPECTED_TOTAL_FINDINGS})" in html_content
301
+ assert "src/exports/tests/test_resume_skip_export.py:35" in html_content
302
+
303
+
304
+ def test_cli_degraded_layout_when_summary_entry_absent(tmp_path: Path) -> None:
305
+ """Should render the timeline and appendix but no scene, table, or rollup markup."""
306
+ run_root = tmp_path / "wf_run_no_summary"
307
+ journal_destination = _copy_run_tree_without_summary_entry(run_root)
308
+
309
+ out_path = tmp_path / "report-degraded.html"
310
+ completed = _render_cli(journal_destination, out_path)
311
+
312
+ assert completed.returncode == 0, f"CLI failed:\n{completed.stderr}"
313
+ html_content = out_path.read_text(encoding="utf-8")
314
+
315
+ assert "PR #211 Convergence Summary" in html_content
316
+ assert 'class="timeline"' not in html_content
317
+ assert "distinct findings across 4 rounds" in html_content
318
+ assert '<details class="appendix"' in html_content
319
+ assert 'class="pf-grid"' not in html_content
320
+ assert 'class="issue-table"' not in html_content
321
+ assert 'class="rollup"' not in html_content
322
+ assert 'class="pr-summary"' not in html_content
323
+
324
+
325
+ def test_cli_injects_summary_from_file_bypassing_transcripts(tmp_path: Path) -> None:
326
+ """Should render the full summary body from --summary-file when no summary transcript exists."""
327
+ run_root = tmp_path / "wf_run_inject"
328
+ journal_destination = _copy_run_tree_without_summary_entry(run_root)
329
+
330
+ summary = {
331
+ "prProblem": "PhotoSync stopped backing up photos after an account switch.",
332
+ "prFix": "It re-checks the account on each backup, so a switch never halts backups.",
333
+ "problemScenes": [],
334
+ "fixScenes": [],
335
+ "verdictLine": "Converged in 4 rounds; every class is fixed.",
336
+ "issueClasses": [
337
+ {
338
+ "plainName": "An injected class the transcript never carried",
339
+ "count": 2,
340
+ "severity": "P1",
341
+ "category": "bug",
342
+ "status": "fixed",
343
+ "cause": "A concrete grounded cause sentence.",
344
+ "medium": "text",
345
+ "beforeLines": [],
346
+ "afterLines": [],
347
+ }
348
+ ],
349
+ }
350
+ summary_path = tmp_path / "summary.json"
351
+ summary_path.write_text(json.dumps(summary), encoding="utf-8")
352
+
353
+ out_path = tmp_path / "report-injected.html"
354
+ render_script = Path(__file__).resolve().parent / "render_report.py"
355
+ completed = subprocess.run(
126
356
  [
127
357
  sys.executable,
128
358
  str(render_script),
129
359
  "--journal",
130
- str(FIXTURE_JOURNAL),
360
+ str(journal_destination),
131
361
  "--out",
132
362
  str(out_path),
133
363
  "--pr",
@@ -136,14 +366,29 @@ def test_html_contains_no_hedging_words(tmp_path: Path) -> None:
136
366
  "7c2f420c4d5b7c83aa47f93d99a0f1420e3373c4",
137
367
  "--rounds",
138
368
  "4",
139
- "--repo",
140
- ".",
369
+ "--summary-file",
370
+ str(summary_path),
141
371
  ],
142
372
  capture_output=True,
143
373
  text=True,
144
- check=True,
145
374
  )
146
375
 
376
+ assert completed.returncode == 0, f"CLI failed:\n{completed.stderr}"
377
+ html_content = out_path.read_text(encoding="utf-8")
378
+ assert 'class="verdict"' in html_content
379
+ assert "Converged in 4 rounds; every class is fixed." in html_content
380
+ assert "An injected class the transcript never carried" in html_content
381
+ assert 'class="pf-grid"' in html_content
382
+ assert f"Raw findings ({EXPECTED_TOTAL_FINDINGS})" in html_content
383
+
384
+
385
+ def test_html_contains_no_hedging_words(tmp_path: Path) -> None:
386
+ """Should produce HTML with no hedging language anywhere in the rendered narrative."""
387
+ out_path = tmp_path / "report-hedge.html"
388
+
389
+ completed = _render_cli(FIXTURE_JOURNAL, out_path)
390
+ assert completed.returncode == 0, f"CLI failed:\n{completed.stderr}"
391
+
147
392
  html_content = out_path.read_text(encoding="utf-8")
148
393
  all_hedging_words = [
149
394
  "could",
@@ -162,141 +407,6 @@ def test_html_contains_no_hedging_words(tmp_path: Path) -> None:
162
407
  )
163
408
 
164
409
 
165
- def _init_git_repo(repo_path: Path) -> None:
166
- """Initialize a git repo with a committed baseline so diffs resolve."""
167
- subprocess.run(
168
- ["git", "-C", str(repo_path), "init"], capture_output=True, check=True
169
- )
170
- subprocess.run(
171
- ["git", "-C", str(repo_path), "config", "user.email", "test@example.com"],
172
- capture_output=True,
173
- check=True,
174
- )
175
- subprocess.run(
176
- ["git", "-C", str(repo_path), "config", "user.name", "Test"],
177
- capture_output=True,
178
- check=True,
179
- )
180
- (repo_path / "README.md").write_text("baseline\n", encoding="utf-8")
181
- subprocess.run(
182
- ["git", "-C", str(repo_path), "add", "."], capture_output=True, check=True
183
- )
184
- subprocess.run(
185
- ["git", "-C", str(repo_path), "commit", "-m", "baseline"],
186
- capture_output=True,
187
- check=True,
188
- )
189
-
190
-
191
- def _resolve_head(repo_path: Path) -> str:
192
- """Return the current HEAD sha of the repo."""
193
- completed = subprocess.run(
194
- ["git", "-C", str(repo_path), "rev-parse", "HEAD"],
195
- capture_output=True,
196
- text=True,
197
- check=True,
198
- )
199
- return completed.stdout.strip()
200
-
201
-
202
- def test_count_tests_added_does_not_double_count_new_file(tmp_path: Path) -> None:
203
- """Should count a new test file with two test functions as exactly two."""
204
- repo_path = tmp_path / "repo"
205
- repo_path.mkdir()
206
- _init_git_repo(repo_path)
207
- base_sha = _resolve_head(repo_path)
208
-
209
- new_test_file = repo_path / "test_feature.py"
210
- new_test_file.write_text(
211
- "def test_one() -> None:\n"
212
- " assert True\n"
213
- "\n"
214
- "def test_two() -> None:\n"
215
- " assert True\n",
216
- encoding="utf-8",
217
- )
218
- subprocess.run(
219
- ["git", "-C", str(repo_path), "add", "."], capture_output=True, check=True
220
- )
221
- subprocess.run(
222
- ["git", "-C", str(repo_path), "commit", "-m", "add tests"],
223
- capture_output=True,
224
- check=True,
225
- )
226
- new_sha = _resolve_head(repo_path)
227
-
228
- test_count = render_report._count_tests_added(base_sha, new_sha, repo_path)
229
-
230
- assert test_count == 2, f"Expected 2 added test definitions, got {test_count}"
231
-
232
-
233
- def test_count_tests_added_counts_nested_test_directory(tmp_path: Path) -> None:
234
- """Should count test functions added under a nested src/<pkg>/tests/ layout."""
235
- repo_path = tmp_path / "repo"
236
- repo_path.mkdir()
237
- _init_git_repo(repo_path)
238
- base_sha = _resolve_head(repo_path)
239
-
240
- nested_test_file = repo_path / "src" / "exports" / "tests" / "test_feature.py"
241
- nested_test_file.parent.mkdir(parents=True)
242
- nested_test_file.write_text(
243
- "def test_one() -> None:\n"
244
- " assert True\n"
245
- "\n"
246
- "def test_two() -> None:\n"
247
- " assert True\n",
248
- encoding="utf-8",
249
- )
250
- subprocess.run(
251
- ["git", "-C", str(repo_path), "add", "."], capture_output=True, check=True
252
- )
253
- subprocess.run(
254
- ["git", "-C", str(repo_path), "commit", "-m", "add nested tests"],
255
- capture_output=True,
256
- check=True,
257
- )
258
- new_sha = _resolve_head(repo_path)
259
-
260
- test_count = render_report._count_tests_added(base_sha, new_sha, repo_path)
261
-
262
- assert test_count == 2, (
263
- f"Expected 2 added test definitions in nested dir, got {test_count}"
264
- )
265
-
266
-
267
- def test_count_tests_added_counts_should_functions(tmp_path: Path) -> None:
268
- """Should count pytest should_* functions, not only def test functions."""
269
- repo_path = tmp_path / "repo"
270
- repo_path.mkdir()
271
- _init_git_repo(repo_path)
272
- base_sha = _resolve_head(repo_path)
273
-
274
- new_test_file = repo_path / "test_behavior.py"
275
- new_test_file.write_text(
276
- "def should_validate_order() -> None:\n"
277
- " assert True\n"
278
- "\n"
279
- "def test_explicit() -> None:\n"
280
- " assert True\n",
281
- encoding="utf-8",
282
- )
283
- subprocess.run(
284
- ["git", "-C", str(repo_path), "add", "."], capture_output=True, check=True
285
- )
286
- subprocess.run(
287
- ["git", "-C", str(repo_path), "commit", "-m", "add should and test"],
288
- capture_output=True,
289
- check=True,
290
- )
291
- new_sha = _resolve_head(repo_path)
292
-
293
- test_count = render_report._count_tests_added(base_sha, new_sha, repo_path)
294
-
295
- assert test_count == 2, (
296
- f"Expected 2 added definitions (should_ + test), got {test_count}"
297
- )
298
-
299
-
300
410
  def test_extract_structured_output_returns_last_tool_input(tmp_path: Path) -> None:
301
411
  """Should return the input of the last StructuredOutput tool_use in the transcript."""
302
412
  transcript_path = tmp_path / "agent-stream.jsonl"
@@ -326,7 +436,9 @@ def test_extract_structured_output_returns_last_tool_input(tmp_path: Path) -> No
326
436
  }
327
437
  }
328
438
  )
329
- transcript_path.write_text(earlier_line + "\n" + later_line + "\n", encoding="utf-8")
439
+ transcript_path.write_text(
440
+ earlier_line + "\n" + later_line + "\n", encoding="utf-8"
441
+ )
330
442
 
331
443
  extracted = render_report._extract_structured_output(transcript_path)
332
444
 
@@ -342,32 +454,14 @@ def test_extract_structured_output_returns_none_on_missing_file(tmp_path: Path)
342
454
  assert extracted is None
343
455
 
344
456
 
345
- def test_render_fix_block_falls_back_when_sha_empty() -> None:
346
- """Should not claim a commit when the fix record has an empty new sha."""
347
- finding = render_report.RawFinding(
348
- file="src/exports/writer.py",
349
- line=10,
350
- severity="P2",
351
- title="example finding",
352
- detail="example detail",
353
- round_number=2,
354
- sha="abc",
457
+ def test_fix_record_carries_summary_text() -> None:
458
+ """Should read the fix agent's summary field into the FixRecord."""
459
+ fix_record = render_report._parse_fix_record(
460
+ {"newSha": "abcd1234", "pushed": True, "summary": "renamed and annotated"},
461
+ base_sha="base",
355
462
  )
356
- fix_by_round = {
357
- 2: render_report.FixRecord(
358
- new_sha="",
359
- pushed=False,
360
- resolved_without_commit=False,
361
- round_number=2,
362
- base_sha="base",
363
- )
364
- }
365
463
 
366
- fix_html = render_report._render_fix_block(finding, fix_by_round)
367
-
368
- assert "<code></code>" not in fix_html
369
- assert "fix commit" not in fix_html
370
- assert "resolved during convergence" in fix_html
464
+ assert fix_record.summary == "renamed and annotated"
371
465
 
372
466
 
373
467
  def _write_structured_output_transcript(
@@ -423,7 +517,10 @@ def test_base_sha_resets_each_round_when_prior_fix_transcript_missing(
423
517
  {"label": render_report.LABEL_PREFIX_FIX + "copilot", "agentId": "missing-fix"},
424
518
  {"label": render_report.LABEL_RESOLVE_HEAD, "agentId": "round-two-resolve"},
425
519
  {"label": render_report.LABEL_COPILOT_GATE, "agentId": round_two_gate_id},
426
- {"label": render_report.LABEL_PREFIX_FIX + "copilot", "agentId": round_two_fix_id},
520
+ {
521
+ "label": render_report.LABEL_PREFIX_FIX + "copilot",
522
+ "agentId": round_two_fix_id,
523
+ },
427
524
  ]
428
525
 
429
526
  _all_findings, fix_by_round = render_report._parse_progress_entries(
@@ -436,26 +533,173 @@ def test_base_sha_resets_each_round_when_prior_fix_transcript_missing(
436
533
  )
437
534
 
438
535
 
439
- def test_robustness_with_missing_transcripts(tmp_path: Path) -> None:
440
- """Should exit 0 and render zero finding cards when no agent transcripts exist."""
441
- run_root = tmp_path / "wf_run"
442
- journal_destination = run_root / "workflows" / FIXTURE_JOURNAL.name
443
- journal_destination.parent.mkdir(parents=True)
444
- shutil.copy(FIXTURE_JOURNAL, journal_destination)
536
+ def _render_cli_with_summary_file(
537
+ journal_path: Path, out_path: Path, summary_path: Path
538
+ ) -> subprocess.CompletedProcess[str]:
539
+ """Run the render CLI with an injected --summary-file and return the process."""
540
+ render_script = Path(__file__).resolve().parent / "render_report.py"
541
+ return subprocess.run(
542
+ [
543
+ sys.executable,
544
+ str(render_script),
545
+ "--journal",
546
+ str(journal_path),
547
+ "--out",
548
+ str(out_path),
549
+ "--pr",
550
+ "example-owner/example-repo#211",
551
+ "--final-sha",
552
+ "7c2f420c4d5b7c83aa47f93d99a0f1420e3373c4",
553
+ "--rounds",
554
+ "4",
555
+ "--summary-file",
556
+ str(summary_path),
557
+ ],
558
+ capture_output=True,
559
+ text=True,
560
+ )
445
561
 
446
- run_id = FIXTURE_JOURNAL.stem
447
- empty_agents_dir = run_root / "subagents" / "workflows" / run_id
448
- empty_agents_dir.mkdir(parents=True)
449
562
 
450
- out_path = tmp_path / "report-robust.html"
451
- render_script = Path(__file__).resolve().parent / "render_report.py"
563
+ def test_cli_renders_when_issue_class_count_is_null(tmp_path: Path) -> None:
564
+ """Should exit 0 and show a zero count when an issue class carries count: null."""
565
+ run_root = tmp_path / "wf_run_null_count"
566
+ journal_destination = _copy_run_tree_without_summary_entry(run_root)
567
+
568
+ summary = {
569
+ "prProblem": "A problem.",
570
+ "prFix": "A fix.",
571
+ "problemScenes": [],
572
+ "fixScenes": [],
573
+ "verdictLine": "Converged.",
574
+ "issueClasses": [
575
+ {
576
+ "plainName": "A class with a null count",
577
+ "count": None,
578
+ "severity": "P2",
579
+ "category": "bug",
580
+ "status": "fixed",
581
+ "cause": "A grounded cause.",
582
+ "medium": "text",
583
+ "beforeLines": [],
584
+ "afterLines": [],
585
+ }
586
+ ],
587
+ }
588
+ summary_path = tmp_path / "summary-null-count.json"
589
+ summary_path.write_text(json.dumps(summary), encoding="utf-8")
590
+
591
+ out_path = tmp_path / "report-null-count.html"
592
+ completed = _render_cli_with_summary_file(
593
+ journal_destination, out_path, summary_path
594
+ )
595
+
596
+ assert completed.returncode == 0, f"CLI crashed on null count:\n{completed.stderr}"
597
+ html_content = out_path.read_text(encoding="utf-8")
598
+ assert "A class with a null count" in html_content
599
+ assert "0 findings" in html_content
600
+ assert "&times;0" in html_content
601
+
602
+
603
+ def test_cli_renders_when_issue_class_count_is_non_numeric(tmp_path: Path) -> None:
604
+ """Should exit 0 and show a zero count when an issue class count is a bad string."""
605
+ run_root = tmp_path / "wf_run_bad_count"
606
+ journal_destination = _copy_run_tree_without_summary_entry(run_root)
607
+
608
+ summary = {
609
+ "prProblem": "A problem.",
610
+ "prFix": "A fix.",
611
+ "problemScenes": [],
612
+ "fixScenes": [],
613
+ "verdictLine": "Converged.",
614
+ "issueClasses": [
615
+ {
616
+ "plainName": "A class with a non-numeric count",
617
+ "count": "x",
618
+ "severity": "P2",
619
+ "category": "bug",
620
+ "status": "fixed",
621
+ "cause": "A grounded cause.",
622
+ "medium": "text",
623
+ "beforeLines": [],
624
+ "afterLines": [],
625
+ }
626
+ ],
627
+ }
628
+ summary_path = tmp_path / "summary-bad-count.json"
629
+ summary_path.write_text(json.dumps(summary), encoding="utf-8")
630
+
631
+ out_path = tmp_path / "report-bad-count.html"
632
+ completed = _render_cli_with_summary_file(
633
+ journal_destination, out_path, summary_path
634
+ )
635
+
636
+ assert completed.returncode == 0, (
637
+ f"CLI crashed on non-numeric count:\n{completed.stderr}"
638
+ )
639
+ html_content = out_path.read_text(encoding="utf-8")
640
+ assert "A class with a non-numeric count" in html_content
641
+ assert "0 findings" in html_content
642
+
643
+
644
+ def test_cli_renders_degraded_body_when_summary_is_a_list(tmp_path: Path) -> None:
645
+ """Should render the degraded layout and exit 0 when --summary-file holds a list."""
646
+ run_root = tmp_path / "wf_run_list_summary"
647
+ journal_destination = _copy_run_tree_without_summary_entry(run_root)
648
+
649
+ summary_path = tmp_path / "summary-list.json"
650
+ summary_path.write_text(json.dumps([]), encoding="utf-8")
651
+
652
+ out_path = tmp_path / "report-list-summary.html"
653
+ completed = _render_cli_with_summary_file(
654
+ journal_destination, out_path, summary_path
655
+ )
656
+
657
+ assert completed.returncode == 0, (
658
+ f"CLI crashed on a list summary:\n{completed.stderr}"
659
+ )
660
+ html_content = out_path.read_text(encoding="utf-8")
661
+ assert "distinct findings across 4 rounds" in html_content
662
+ assert 'class="pf-grid"' not in html_content
663
+
452
664
 
665
+ def test_cli_renders_degraded_body_when_summary_is_a_scalar(tmp_path: Path) -> None:
666
+ """Should render the degraded layout and exit 0 when --summary-file holds a scalar."""
667
+ run_root = tmp_path / "wf_run_scalar_summary"
668
+ journal_destination = _copy_run_tree_without_summary_entry(run_root)
669
+
670
+ summary_path = tmp_path / "summary-scalar.json"
671
+ summary_path.write_text(json.dumps(5), encoding="utf-8")
672
+
673
+ out_path = tmp_path / "report-scalar-summary.html"
674
+ completed = _render_cli_with_summary_file(
675
+ journal_destination, out_path, summary_path
676
+ )
677
+
678
+ assert completed.returncode == 0, (
679
+ f"CLI crashed on a scalar summary:\n{completed.stderr}"
680
+ )
681
+ html_content = out_path.read_text(encoding="utf-8")
682
+ assert "distinct findings across 4 rounds" in html_content
683
+ assert 'class="pf-grid"' not in html_content
684
+
685
+
686
+ def test_is_summary_structurally_valid_false_for_non_dict_summary() -> None:
687
+ """Should return False for a list, string, or scalar summary, never raising."""
688
+ assert render_report._is_summary_structurally_valid([]) is False
689
+ assert render_report._is_summary_structurally_valid("str") is False
690
+ assert render_report._is_summary_structurally_valid(5) is False
691
+
692
+
693
+ def test_cli_rejects_orphaned_repo_argument(tmp_path: Path) -> None:
694
+ """Should reject --repo with a usage error, proving the flag is no longer declared."""
695
+ render_script = Path(__file__).resolve().parent / "render_report.py"
696
+ out_path = tmp_path / "report-repo-rejected.html"
453
697
  completed = subprocess.run(
454
698
  [
455
699
  sys.executable,
456
700
  str(render_script),
457
701
  "--journal",
458
- str(journal_destination),
702
+ str(FIXTURE_JOURNAL),
459
703
  "--out",
460
704
  str(out_path),
461
705
  "--pr",
@@ -471,14 +715,29 @@ def test_robustness_with_missing_transcripts(tmp_path: Path) -> None:
471
715
  text=True,
472
716
  )
473
717
 
718
+ assert completed.returncode != 0
719
+ assert "unrecognized arguments: --repo" in completed.stderr
720
+
721
+
722
+ def test_robustness_with_missing_transcripts(tmp_path: Path) -> None:
723
+ """Should exit 0 and render the timeline and appendix when no transcripts exist."""
724
+ run_root = tmp_path / "wf_run"
725
+ journal_destination = run_root / "workflows" / FIXTURE_JOURNAL.name
726
+ journal_destination.parent.mkdir(parents=True)
727
+ shutil.copy(FIXTURE_JOURNAL, journal_destination)
728
+
729
+ run_id = FIXTURE_JOURNAL.stem
730
+ empty_agents_dir = run_root / "subagents" / "workflows" / run_id
731
+ empty_agents_dir.mkdir(parents=True)
732
+
733
+ out_path = tmp_path / "report-robust.html"
734
+ completed = _render_cli(journal_destination, out_path)
735
+
474
736
  assert completed.returncode == 0, (
475
737
  f"Render failed despite missing transcripts:\n{completed.stderr}"
476
738
  )
477
739
 
478
740
  html_content = out_path.read_text(encoding="utf-8")
479
- assert "PR #211 Convergence Insights" in html_content
480
-
481
- finding_card_count = html_content.count('class="bug-card')
482
- assert finding_card_count == 0, (
483
- f"Missing transcripts yielded findings: expected 0 cards, got {finding_card_count}"
484
- )
741
+ assert "PR #211 Convergence Summary" in html_content
742
+ assert 'class="timeline"' not in html_content
743
+ assert 'class="pf-grid"' not in html_content