evidence-cli 0.6.1__tar.gz → 0.7.1__tar.gz

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 (136) hide show
  1. evidence_cli-0.7.1/.devague/current +1 -0
  2. evidence_cli-0.7.1/.devague/current_plan +1 -0
  3. evidence_cli-0.7.1/.devague/deliveries/evidence-cli-run-check-and-answer-check.json +519 -0
  4. evidence_cli-0.7.1/.devague/frames/evidence-cli-run-check-and-answer-check.json +852 -0
  5. evidence_cli-0.7.1/.devague/plans/evidence-cli-run-check-and-answer-check.json +574 -0
  6. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/CHANGELOG.md +104 -0
  7. evidence_cli-0.7.1/CLAUDE.md +245 -0
  8. evidence_cli-0.7.1/PKG-INFO +92 -0
  9. evidence_cli-0.7.1/README.md +75 -0
  10. evidence_cli-0.7.1/docs/deliveries/2026-11-18-evidence-cli-run-check-and-answer-check.md +110 -0
  11. evidence_cli-0.7.1/docs/library.md +53 -0
  12. evidence_cli-0.7.1/docs/notes/2026-11-18-live-test.md +71 -0
  13. evidence_cli-0.7.1/docs/plans/2026-11-18-evidence-cli-run-check-and-answer-check-split.md +168 -0
  14. evidence_cli-0.7.1/docs/plans/2026-11-18-evidence-cli-run-check-and-answer-check.md +114 -0
  15. evidence_cli-0.7.1/docs/specs/2026-11-18-evidence-cli-run-check-and-answer-check.md +108 -0
  16. evidence_cli-0.7.1/evidence/accuracy/__init__.py +1 -0
  17. evidence_cli-0.7.1/evidence/accuracy/check.py +1098 -0
  18. evidence_cli-0.7.1/evidence/accuracy/counts.py +61 -0
  19. evidence_cli-0.7.1/evidence/accuracy/matcher.py +81 -0
  20. evidence_cli-0.7.1/evidence/accuracy/vectors.py +15 -0
  21. evidence_cli-0.7.1/evidence/answer/__init__.py +54 -0
  22. evidence_cli-0.7.1/evidence/answer/grade.py +80 -0
  23. evidence_cli-0.7.1/evidence/answer/schema.py +85 -0
  24. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/__init__.py +9 -5
  25. evidence_cli-0.7.1/evidence/cli/_commands/check.py +470 -0
  26. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/_commands/cli.py +4 -4
  27. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/_commands/doctor.py +2 -2
  28. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/_commands/explain.py +2 -2
  29. evidence_cli-0.7.1/evidence/cli/_commands/grade.py +118 -0
  30. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/_commands/learn.py +33 -15
  31. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/_commands/overview.py +1 -1
  32. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/_commands/whoami.py +1 -1
  33. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/_errors.py +1 -1
  34. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/_output.py +20 -3
  35. evidence_cli-0.7.1/evidence/data/accuracy_defaults.json +112 -0
  36. evidence_cli-0.7.1/evidence/data/expected_values.schema.json +47 -0
  37. evidence_cli-0.7.1/evidence/data/matcher_vectors.json +2818 -0
  38. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/explain/__init__.py +1 -1
  39. evidence_cli-0.7.1/evidence/explain/catalog.py +172 -0
  40. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/pyproject.toml +1 -1
  41. evidence_cli-0.7.1/tests/fixtures/accuracy/final_message_w999.json +48 -0
  42. evidence_cli-0.7.1/tests/fixtures/accuracy/finish_summary_pyproject.json +24 -0
  43. evidence_cli-0.7.1/tests/fixtures/accuracy/no_purpose_calls.json +29 -0
  44. evidence_cli-0.7.1/tests/fixtures/accuracy/prior_run_ids.json +217 -0
  45. evidence_cli-0.7.1/tests/fixtures/accuracy/quote_absent_example_markers.json +67 -0
  46. evidence_cli-0.7.1/tests/fixtures/accuracy/senses_remember_success.json +13 -0
  47. evidence_cli-0.7.1/tests/fixtures/answer/count-tests/answer.txt +1 -0
  48. evidence_cli-0.7.1/tests/fixtures/answer/count-tests/expected.json +1 -0
  49. evidence_cli-0.7.1/tests/fixtures/answer/count-tests/forbidden.txt +1 -0
  50. evidence_cli-0.7.1/tests/fixtures/answer/count-tests/wrong.txt +1 -0
  51. evidence_cli-0.7.1/tests/fixtures/answer/find-file/answer.txt +1 -0
  52. evidence_cli-0.7.1/tests/fixtures/answer/find-file/expected.json +1 -0
  53. evidence_cli-0.7.1/tests/fixtures/answer/find-file/wrong.txt +1 -0
  54. evidence_cli-0.7.1/tests/fixtures/answer/stack-version/answer.txt +1 -0
  55. evidence_cli-0.7.1/tests/fixtures/answer/stack-version/expected.json +1 -0
  56. evidence_cli-0.7.1/tests/fixtures/answer/stack-version/wrong.txt +1 -0
  57. evidence_cli-0.7.1/tests/test_answer.py +324 -0
  58. evidence_cli-0.7.1/tests/test_boundaries.py +201 -0
  59. evidence_cli-0.7.1/tests/test_check.py +600 -0
  60. evidence_cli-0.7.1/tests/test_check_command.py +521 -0
  61. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/tests/test_cli.py +19 -2
  62. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/tests/test_cli_introspection.py +3 -3
  63. evidence_cli-0.7.1/tests/test_counts.py +52 -0
  64. evidence_cli-0.7.1/tests/test_end_to_end.py +394 -0
  65. evidence_cli-0.7.1/tests/test_grade_command.py +160 -0
  66. evidence_cli-0.7.1/tests/test_matcher.py +50 -0
  67. evidence_cli-0.7.1/tests/test_public_names.py +55 -0
  68. evidence_cli-0.7.1/tests/test_release_docs.py +70 -0
  69. evidence_cli-0.7.1/tests/test_replays.py +187 -0
  70. evidence_cli-0.7.1/tests/test_runs.py +326 -0
  71. evidence_cli-0.7.1/tests/test_wording.py +145 -0
  72. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/uv.lock +1 -1
  73. evidence_cli-0.6.1/CLAUDE.md +0 -28
  74. evidence_cli-0.6.1/PKG-INFO +0 -76
  75. evidence_cli-0.6.1/README.md +0 -59
  76. evidence_cli-0.6.1/evidence/explain/catalog.py +0 -130
  77. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/agent-config/SKILL.md +0 -0
  78. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/agent-config/data/backend-fingerprints.yaml +0 -0
  79. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/agent-config/scripts/show.sh +0 -0
  80. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/ask-colleague/SKILL.md +0 -0
  81. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/ask-colleague/prompts/explore.md +0 -0
  82. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/ask-colleague/prompts/review.md +0 -0
  83. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/ask-colleague/prompts/write.md +0 -0
  84. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/ask-colleague/scripts/ask-colleague.sh +0 -0
  85. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/assign-to-workforce/SKILL.md +0 -0
  86. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/assign-to-workforce/scripts/assign-to-workforce.sh +0 -0
  87. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/challenge/SKILL.md +0 -0
  88. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/cicd/SKILL.md +0 -0
  89. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/cicd/scripts/_resolve-nick.sh +0 -0
  90. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/cicd/scripts/portability-lint.sh +0 -0
  91. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/cicd/scripts/pr-reply.sh +0 -0
  92. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/cicd/scripts/pr-status.sh +0 -0
  93. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/cicd/scripts/workflow.sh +0 -0
  94. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/communicate/SKILL.md +0 -0
  95. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/communicate/scripts/fetch-issues.sh +0 -0
  96. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/communicate/scripts/mesh-message.sh +0 -0
  97. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/communicate/scripts/post-comment.sh +0 -0
  98. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/communicate/scripts/post-issue.sh +0 -0
  99. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/communicate/scripts/templates/skill-new-brief.md +0 -0
  100. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/communicate/scripts/templates/skill-update-brief.md +0 -0
  101. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/deviate/SKILL.md +0 -0
  102. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/doc-test-alignment/SKILL.md +0 -0
  103. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/doc-test-alignment/scripts/check.sh +0 -0
  104. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/pypi-maintainer/SKILL.md +0 -0
  105. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/pypi-maintainer/scripts/switch-source.sh +0 -0
  106. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/recall/SKILL.md +0 -0
  107. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/recall/scripts/recall.sh +0 -0
  108. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/remember/SKILL.md +0 -0
  109. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/remember/scripts/remember.sh +0 -0
  110. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/run-tests/SKILL.md +0 -0
  111. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/run-tests/scripts/test.sh +0 -0
  112. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/scope/SKILL.md +0 -0
  113. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/sonarclaude/SKILL.md +0 -0
  114. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/sonarclaude/scripts/sonar.sh +0 -0
  115. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/spec-to-plan/SKILL.md +0 -0
  116. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/spec-to-plan/scripts/spec-to-plan.sh +0 -0
  117. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/summarize-delivery/SKILL.md +0 -0
  118. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/think/SKILL.md +0 -0
  119. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/think/scripts/think.sh +0 -0
  120. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/version-bump/SKILL.md +0 -0
  121. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills/version-bump/scripts/bump.py +0 -0
  122. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.claude/skills.local.yaml.example +0 -0
  123. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.flake8 +0 -0
  124. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.github/workflows/publish.yml +0 -0
  125. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.github/workflows/tests.yml +0 -0
  126. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.gitignore +0 -0
  127. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/.markdownlint-cli2.yaml +0 -0
  128. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/AGENTS.colleague.md +0 -0
  129. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/LICENSE +0 -0
  130. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/culture.yaml +0 -0
  131. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/docs/skill-sources.md +0 -0
  132. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/__init__.py +0 -0
  133. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/__main__.py +0 -0
  134. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/evidence/cli/_commands/__init__.py +0 -0
  135. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/sonar-project.properties +0 -0
  136. {evidence_cli-0.6.1 → evidence_cli-0.7.1}/tests/__init__.py +0 -0
@@ -0,0 +1 @@
1
+ evidence-cli-run-check-and-answer-check
@@ -0,0 +1 @@
1
+ evidence-cli-run-check-and-answer-check
@@ -0,0 +1,519 @@
1
+ {
2
+ "plan_slug": "evidence-cli-run-check-and-answer-check",
3
+ "schema_version": 2,
4
+ "created": "2026-11-18T15:46:15Z",
5
+ "updated": "2026-11-18T15:52:27Z",
6
+ "deviations": [
7
+ {
8
+ "id": "d1",
9
+ "what": "The moved checker was changed to refuse export files linked from outside the export, although the plan said to copy it keeping its results",
10
+ "task_ref": "t2",
11
+ "reason": "the t7 boundary test found facts.json followed a link out of the export; the operator said on 2026-11-18 to fix it in PR 4",
12
+ "affects": [],
13
+ "origin": "user",
14
+ "status": "approved",
15
+ "classification": "acceptable",
16
+ "seq": 21
17
+ },
18
+ {
19
+ "id": "d2",
20
+ "what": "The printed program name was changed from evidence-cli to evidence across usage, hints, learn and explain; no task covered it",
21
+ "task_ref": "t6",
22
+ "reason": "the installed command is evidence; the operator said on 2026-11-18 to fix the name in PR 4",
23
+ "affects": [],
24
+ "origin": "user",
25
+ "status": "approved",
26
+ "classification": "acceptable",
27
+ "seq": 22
28
+ },
29
+ {
30
+ "id": "d3",
31
+ "what": "Fourteen code-review findings were fixed on the PR after the waves, changing eleven behaviours of the moved checker and the answer check's text comparison; I started the fixes before asking",
32
+ "task_ref": "t2",
33
+ "reason": "the operator ran a code review on PR 4; every finding reproduced; spec claim c16 was amended and confirmed afterwards",
34
+ "affects": [],
35
+ "origin": "llm",
36
+ "status": "approved",
37
+ "classification": "needs-follow-up",
38
+ "seq": 23
39
+ },
40
+ {
41
+ "id": "d4",
42
+ "what": "Task t5 was started before its wave (it depended only on t4) and was allowed to create tests/fixtures/answer/, outside its owned files; that collided with t2's fixture checksum test, which I rescoped after the merge",
43
+ "task_ref": "t5",
44
+ "reason": "to use waiting time while t2 ran; t5's first criterion needed fixture tasks that did not exist",
45
+ "affects": [],
46
+ "origin": "llm",
47
+ "status": "approved",
48
+ "classification": "acceptable",
49
+ "seq": 24
50
+ },
51
+ {
52
+ "id": "d5",
53
+ "what": "After the waves I changed code myself, outside any task: conclusion counts in the text report, the backed and refused counts found by the live test, template text in learn and explain, a forbidden-value fixture, and SonarCloud fixes by a further agent",
54
+ "task_ref": "t6",
55
+ "reason": "gaps the task agents reported in their own hand-backs, the live test's first finding, and 33 SonarCloud issues",
56
+ "affects": [],
57
+ "origin": "llm",
58
+ "status": "approved",
59
+ "classification": "acceptable",
60
+ "seq": 25
61
+ },
62
+ {
63
+ "id": "d6",
64
+ "what": "The live test ran before task t9 was confirmed, and its third task (a review of the new code) was added to my first description of it",
65
+ "task_ref": "t9",
66
+ "reason": "the operator asked for associate as live tester on worker and cortex; I started the runs while the last build tasks were in progress; the operator confirmed t9 afterwards",
67
+ "affects": [],
68
+ "origin": "llm",
69
+ "status": "approved",
70
+ "classification": "acceptable",
71
+ "seq": 26
72
+ }
73
+ ],
74
+ "evidence": [
75
+ {
76
+ "id": "e1",
77
+ "obligation_ref": "o1",
78
+ "test_ref": "tests/test_check.py::test_each_public_function_is_importable_from_its_module + tests/test_public_names.py::test_the_eight_names_associate_imports_are_listed",
79
+ "behavior_text": "asserts the eight names import from their documented modules and are listed in docs/library.md",
80
+ "contract_text": "each name associate imports is importable under its documented path",
81
+ "evidence_type": "automated",
82
+ "strength": "execution",
83
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main. PARTIAL: the second half of the obligation (the same result as the associate version on every saved test export) has no side-by-side test; it rests on the ported replay tests and on three live exports where the counts equalled associate check's",
84
+ "outcome": "pass",
85
+ "run": {
86
+ "timestamp": "2026-11-18T15:50:00+00:00",
87
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
88
+ },
89
+ "origin": "llm",
90
+ "status": "approved",
91
+ "superseded": false,
92
+ "seq": 1
93
+ },
94
+ {
95
+ "id": "e2",
96
+ "obligation_ref": "o1",
97
+ "test_ref": "docs/notes/2026-11-18-live-test.md (three live exports, counts compared with associate check by script)",
98
+ "behavior_text": "evidence check and associate check printed the same fact counts on three live exports",
99
+ "contract_text": "returns the result the associate version returned",
100
+ "evidence_type": "observation",
101
+ "strength": "fidelity",
102
+ "strength_basis": "one comparison on three exports made on 2026-11-18 at branch commit 024c92b, before eleven behaviour changes from the code review; the fact counts were re-run after those changes and were unchanged",
103
+ "outcome": "pass",
104
+ "run": {
105
+ "timestamp": "2026-11-18T15:50:00+00:00",
106
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
107
+ },
108
+ "origin": "llm",
109
+ "status": "approved",
110
+ "superseded": false,
111
+ "seq": 2
112
+ },
113
+ {
114
+ "id": "e3",
115
+ "obligation_ref": "o2",
116
+ "test_ref": "tests/test_matcher.py::test_vector + tests/test_check.py::test_python_checker_reproduces_every_handback_vector + tests/test_end_to_end.py::test_every_shared_vector_passes_through_the_entry_point_handback",
117
+ "behavior_text": "asserts every matcher case and every hand-back case in the shared vectors returns the expected verdict, in the library and through the entry point",
118
+ "contract_text": "the moved matcher and checker return the expected verdict on every shared vector",
119
+ "evidence_type": "automated",
120
+ "strength": "sensitivity",
121
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main; the task agent saw this test fail under a temporary mutation of the code it guards",
122
+ "outcome": "pass",
123
+ "run": {
124
+ "timestamp": "2026-11-18T15:50:00+00:00",
125
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
126
+ },
127
+ "origin": "llm",
128
+ "status": "approved",
129
+ "superseded": false,
130
+ "seq": 3
131
+ },
132
+ {
133
+ "id": "e4",
134
+ "obligation_ref": "o3",
135
+ "test_ref": "tests/test_check_command.py::test_settings_named_in_output",
136
+ "behavior_text": "asserts the output names the bundled default and its version with no settings file, and names the file when one is given",
137
+ "contract_text": "the output states which settings were used",
138
+ "evidence_type": "automated",
139
+ "strength": "execution",
140
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main",
141
+ "outcome": "pass",
142
+ "run": {
143
+ "timestamp": "2026-11-18T15:50:00+00:00",
144
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
145
+ },
146
+ "origin": "llm",
147
+ "status": "approved",
148
+ "superseded": false,
149
+ "seq": 4
150
+ },
151
+ {
152
+ "id": "e5",
153
+ "obligation_ref": "o4",
154
+ "test_ref": "tests/test_check_command.py::test_per_run_counts_and_sums_equal_hand_count + tests/test_end_to_end.py::test_counts_over_the_fixture_exports_equal_a_hand_count",
155
+ "behavior_text": "asserts per-run counts by kind and source kind and their sums equal literals counted by hand",
156
+ "contract_text": "counts for each run and their sum equal a hand count",
157
+ "evidence_type": "automated",
158
+ "strength": "sensitivity",
159
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main; the task agent saw this test fail under a temporary mutation of the code it guards",
160
+ "outcome": "pass",
161
+ "run": {
162
+ "timestamp": "2026-11-18T15:50:00+00:00",
163
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
164
+ },
165
+ "origin": "llm",
166
+ "status": "approved",
167
+ "superseded": false,
168
+ "seq": 5
169
+ },
170
+ {
171
+ "id": "e6",
172
+ "obligation_ref": "o5",
173
+ "test_ref": "tests/test_check_command.py::test_wrong_report_count_is_a_disagreement_with_both_values",
174
+ "behavior_text": "asserts a report.json with a wrong count is listed with both values",
175
+ "contract_text": "a differing count is listed as a disagreement with both values",
176
+ "evidence_type": "automated",
177
+ "strength": "execution",
178
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main",
179
+ "outcome": "pass",
180
+ "run": {
181
+ "timestamp": "2026-11-18T15:50:00+00:00",
182
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
183
+ },
184
+ "origin": "llm",
185
+ "status": "approved",
186
+ "superseded": false,
187
+ "seq": 6
188
+ },
189
+ {
190
+ "id": "e7",
191
+ "obligation_ref": "o6",
192
+ "test_ref": "tests/test_check_command.py::test_no_facts_run_is_not_applicable_and_adds_nothing + tests/test_check.py::test_an_old_export_is_read_and_its_facts_are_not_applicable",
193
+ "behavior_text": "asserts a run with no facts file, and an old walk, are not applicable and add nothing to the sums",
194
+ "contract_text": "a run with no facts is not applicable and adds nothing",
195
+ "evidence_type": "automated",
196
+ "strength": "execution",
197
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main",
198
+ "outcome": "pass",
199
+ "run": {
200
+ "timestamp": "2026-11-18T15:50:00+00:00",
201
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
202
+ },
203
+ "origin": "llm",
204
+ "status": "approved",
205
+ "superseded": false,
206
+ "seq": 7
207
+ },
208
+ {
209
+ "id": "e8",
210
+ "obligation_ref": "o7",
211
+ "test_ref": "tests/test_end_to_end.py::test_the_flags_file_holds_a_reviewer_flag_with_the_export_path + tests/test_end_to_end.py::test_every_export_file_is_byte_identical_after_both_commands + tests/test_check_command.py::test_flags_file_is_not_written_inside_the_export_when_run_from_inside_it",
212
+ "behavior_text": "asserts a real reviewer flag lands in the file beside the export and every export file keeps its checksum",
213
+ "contract_text": "flags are written beside the run and the export is unchanged",
214
+ "evidence_type": "automated",
215
+ "strength": "sensitivity",
216
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main; the task agent saw this test fail under a temporary mutation of the code it guards",
217
+ "outcome": "pass",
218
+ "run": {
219
+ "timestamp": "2026-11-18T15:50:00+00:00",
220
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
221
+ },
222
+ "origin": "llm",
223
+ "status": "approved",
224
+ "superseded": false,
225
+ "seq": 8
226
+ },
227
+ {
228
+ "id": "e9",
229
+ "obligation_ref": "o8",
230
+ "test_ref": "tests/test_grade_command.py::test_recorded_answer_matches_and_wrong_does_not + tests/test_grade_command.py::test_empty_answer_is_cannot_tell + tests/test_grade_command.py::test_text_lists_found_missing_and_schema_version",
231
+ "behavior_text": "asserts the three results with the items found and missing",
232
+ "contract_text": "the result is matches, does not match or cannot tell with items found and missing",
233
+ "evidence_type": "automated",
234
+ "strength": "execution",
235
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main. The clause about the export's facts file being absent holds by interface: the command takes an answer file and an expected-values file and no export; no test deletes a facts file",
236
+ "outcome": "pass",
237
+ "run": {
238
+ "timestamp": "2026-11-18T15:50:00+00:00",
239
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
240
+ },
241
+ "origin": "llm",
242
+ "status": "approved",
243
+ "superseded": false,
244
+ "seq": 9
245
+ },
246
+ {
247
+ "id": "e10",
248
+ "obligation_ref": "o9",
249
+ "test_ref": "tests/test_grade_command.py::test_two_runs_are_byte_identical + tests/test_end_to_end.py::test_a_second_run_prints_a_byte_identical_grade",
250
+ "behavior_text": "asserts two runs on the same inputs print identical bytes, in text and JSON, in one process and in separate processes",
251
+ "contract_text": "the same inputs give byte-identical results",
252
+ "evidence_type": "automated",
253
+ "strength": "execution",
254
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main",
255
+ "outcome": "pass",
256
+ "run": {
257
+ "timestamp": "2026-11-18T15:50:00+00:00",
258
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
259
+ },
260
+ "origin": "llm",
261
+ "status": "approved",
262
+ "superseded": false,
263
+ "seq": 10
264
+ },
265
+ {
266
+ "id": "e11",
267
+ "obligation_ref": "o10",
268
+ "test_ref": "tests/test_answer.py::test_bad_expected_refused_with_reason + tests/test_grade_command.py::test_refused_expected_file_is_user_error_with_remediation + tests/test_answer.py::test_every_result_names_schema_version",
269
+ "behavior_text": "asserts an unknown key or missing schema version is refused with the reason and every result names the schema version",
270
+ "contract_text": "a bad task file is refused with the reason; results name the schema version",
271
+ "evidence_type": "automated",
272
+ "strength": "execution",
273
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main",
274
+ "outcome": "pass",
275
+ "run": {
276
+ "timestamp": "2026-11-18T15:50:00+00:00",
277
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
278
+ },
279
+ "origin": "llm",
280
+ "status": "approved",
281
+ "superseded": false,
282
+ "seq": 11
283
+ },
284
+ {
285
+ "id": "e12",
286
+ "obligation_ref": "o11",
287
+ "test_ref": "tests/test_wording.py (test_check_text_wording, test_check_json_wording, test_grade_text_wording, test_grade_json_wording)",
288
+ "behavior_text": "asserts both commands' text and JSON carry the line about a complete trail and contain neither quality score nor accuracy",
289
+ "contract_text": "the output names its result, carries the line, and avoids the two phrases",
290
+ "evidence_type": "automated",
291
+ "strength": "execution",
292
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main. The fixtures behind these tests are three small exports; the unwritable-location message and the --settings text path are not covered by the wording test",
293
+ "outcome": "pass",
294
+ "run": {
295
+ "timestamp": "2026-11-18T15:50:00+00:00",
296
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
297
+ },
298
+ "origin": "llm",
299
+ "status": "approved",
300
+ "superseded": false,
301
+ "seq": 12
302
+ },
303
+ {
304
+ "id": "e13",
305
+ "obligation_ref": "o12",
306
+ "test_ref": "tests/test_matcher.py::test_vectors_are_read_through_the_packaged_function + tests/test_boundaries.py::test_the_built_wheel_carries_the_data_files_and_requires_nothing + tests/test_end_to_end.py::test_the_bundled_data_files_are_byte_copies_of_associate_at_266dad3",
307
+ "behavior_text": "asserts the vectors are read from the package, ship in the built wheel, and match a pinned checksum",
308
+ "contract_text": "the shared vector files are readable from the installed package and equal the suite's",
309
+ "evidence_type": "automated",
310
+ "strength": "sensitivity",
311
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main; the task agent saw this test fail under a temporary mutation of the code it guards",
312
+ "outcome": "pass",
313
+ "run": {
314
+ "timestamp": "2026-11-18T15:50:00+00:00",
315
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
316
+ },
317
+ "origin": "llm",
318
+ "status": "approved",
319
+ "superseded": false,
320
+ "seq": 13
321
+ },
322
+ {
323
+ "id": "e14",
324
+ "obligation_ref": "o13",
325
+ "test_ref": "tests/test_release_docs.py::test_claude_md_is_no_longer_a_seed + tests/test_release_docs.py::test_claude_md_states_the_two_results_names_and_dependency_rule",
326
+ "behavior_text": "asserts CLAUDE.md has no seed re-init instruction and states the two results, the public names and the no-dependency rule",
327
+ "contract_text": "CLAUDE.md states the two results, the names and the rule",
328
+ "evidence_type": "automated",
329
+ "strength": "execution",
330
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main. The test checks that phrases are present; the main agent also read the file once in full on 2026-11-18",
331
+ "outcome": "pass",
332
+ "run": {
333
+ "timestamp": "2026-11-18T15:50:00+00:00",
334
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
335
+ },
336
+ "origin": "llm",
337
+ "status": "approved",
338
+ "superseded": false,
339
+ "seq": 14
340
+ },
341
+ {
342
+ "id": "e15",
343
+ "obligation_ref": "o14",
344
+ "test_ref": "tests/test_boundaries.py::test_pyproject_declares_no_runtime_dependency + tests/test_boundaries.py::test_no_module_under_evidence_imports_associate_a_network_library_or_subprocess + tests/test_boundaries.py::test_the_built_wheel_carries_the_data_files_and_requires_nothing",
345
+ "behavior_text": "asserts an empty dependency list in pyproject and in the built wheel's metadata, and no network or subprocess import under evidence/",
346
+ "contract_text": "the dependency list is empty and no module imports a network or subprocess library",
347
+ "evidence_type": "automated",
348
+ "strength": "sensitivity",
349
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main; the task agent saw this test fail under a temporary mutation of the code it guards",
350
+ "outcome": "pass",
351
+ "run": {
352
+ "timestamp": "2026-11-18T15:50:00+00:00",
353
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
354
+ },
355
+ "origin": "llm",
356
+ "status": "approved",
357
+ "superseded": false,
358
+ "seq": 15
359
+ },
360
+ {
361
+ "id": "e16",
362
+ "obligation_ref": "o15",
363
+ "test_ref": "tests/test_check.py::test_no_module_under_evidence_imports_associate + tests/test_replays.py::test_the_replays_leave_every_fixture_file_unchanged",
364
+ "behavior_text": "asserts no module imports associate and fixture checksums are equal before and after the replays",
365
+ "contract_text": "no module imports associate; fixtures are unchanged",
366
+ "evidence_type": "automated",
367
+ "strength": "execution",
368
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main. The fixture comparison brackets the replay tests, not the whole session",
369
+ "outcome": "pass",
370
+ "run": {
371
+ "timestamp": "2026-11-18T15:50:00+00:00",
372
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
373
+ },
374
+ "origin": "llm",
375
+ "status": "approved",
376
+ "superseded": false,
377
+ "seq": 16
378
+ },
379
+ {
380
+ "id": "e17",
381
+ "obligation_ref": "o16",
382
+ "test_ref": "tests/test_check.py::test_a_newer_walk_format_is_unsupported_with_the_version_found_and_no_counts + tests/test_check_command.py::test_newer_walk_format_is_unsupported_and_adds_nothing + tests/test_check.py::test_a_newer_format_export_is_unsupported_before_its_facts_file_is_read",
383
+ "behavior_text": "asserts a newer format gives an unsupported result naming the version, with no counts, before the facts file is read",
384
+ "contract_text": "a newer format is reported unsupported with the version found and no counts",
385
+ "evidence_type": "automated",
386
+ "strength": "execution",
387
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main; these tests failed before the behaviour was added",
388
+ "outcome": "pass",
389
+ "run": {
390
+ "timestamp": "2026-11-18T15:50:00+00:00",
391
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
392
+ },
393
+ "origin": "llm",
394
+ "status": "approved",
395
+ "superseded": false,
396
+ "seq": 17
397
+ },
398
+ {
399
+ "id": "e18",
400
+ "obligation_ref": "o17",
401
+ "test_ref": "tests/test_answer.py::test_number_whole_token + tests/test_answer.py::test_identifier_whole_token",
402
+ "behavior_text": "asserts a number or identifier does not match inside a longer one and does match next to punctuation",
403
+ "contract_text": "a longer number or identifier containing the expected one does not match",
404
+ "evidence_type": "automated",
405
+ "strength": "execution",
406
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main; the main agent also probed 22 further cases by hand on 2026-11-18",
407
+ "outcome": "pass",
408
+ "run": {
409
+ "timestamp": "2026-11-18T15:50:00+00:00",
410
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
411
+ },
412
+ "origin": "llm",
413
+ "status": "approved",
414
+ "superseded": false,
415
+ "seq": 18
416
+ },
417
+ {
418
+ "id": "e19",
419
+ "obligation_ref": "o18",
420
+ "test_ref": "tests/test_answer.py::test_empty_answer_cannot_tell + tests/test_answer.py::test_no_items_cannot_tell + tests/test_answer.py::test_cause_only_for_cannot_tell + tests/test_answer.py::test_missing_required_is_does_not_match_and_named",
421
+ "behavior_text": "asserts cannot tell with its cause only for an empty answer or no items, and does not match for a missing item",
422
+ "contract_text": "cannot tell only for the two causes; a missing item is does not match",
423
+ "evidence_type": "automated",
424
+ "strength": "execution",
425
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main",
426
+ "outcome": "pass",
427
+ "run": {
428
+ "timestamp": "2026-11-18T15:50:00+00:00",
429
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
430
+ },
431
+ "origin": "llm",
432
+ "status": "approved",
433
+ "superseded": false,
434
+ "seq": 19
435
+ },
436
+ {
437
+ "id": "e20",
438
+ "obligation_ref": "o19",
439
+ "test_ref": "tests/test_check_command.py::test_unwritable_location_prints_report_and_exits_env_error + tests/test_check_command.py::test_flags_dir_and_no_flags_exit_zero + tests/test_check_command.py::test_flags_file_failure_under_json_is_one_json_error_on_stderr",
440
+ "behavior_text": "asserts the report is printed, the message says the file was not written, the exit code is the environment error code, and the two options exit 0",
441
+ "contract_text": "an unwritable location prints the report and exits with an error code; the options exit 0",
442
+ "evidence_type": "automated",
443
+ "strength": "execution",
444
+ "strength_basis": "ran in the full suite (411 passed) at the merge commit on main. The unwritable case is a path that collides with a file; a directory without write permission was not tested",
445
+ "outcome": "pass",
446
+ "run": {
447
+ "timestamp": "2026-11-18T15:50:00+00:00",
448
+ "commit": "f52599de2bc6dbc10fc8c0a932727137d547e34e"
449
+ },
450
+ "origin": "llm",
451
+ "status": "approved",
452
+ "superseded": false,
453
+ "seq": 20
454
+ }
455
+ ],
456
+ "deltas": [
457
+ {
458
+ "id": "b1",
459
+ "kind": "amended",
460
+ "behavior_text": "eleven behaviours of the moved checker differ from associate at 266dad3 (listed in the 0.7.0 changelog): the export's own files are not read through a link pointing outside it, a cut or over-deep walk line is skipped, the walk format is tested before anything else is read, a newer-format child or prior run is not loaded and is listed, a marker without a positive integer version is unsupported, a relative child path resolves against the export, and load_export returns the raw marker value when it is not a valid version",
461
+ "caused_by": [
462
+ "c11"
463
+ ],
464
+ "evidence_refs": [],
465
+ "origin": "llm",
466
+ "status": "approved",
467
+ "superseded": false
468
+ },
469
+ {
470
+ "id": "b2",
471
+ "kind": "added",
472
+ "behavior_text": "the run check reports each run's stated conclusions as backed, left open or refused by the re-check, per run and summed",
473
+ "caused_by": [
474
+ "c13"
475
+ ],
476
+ "evidence_refs": [],
477
+ "origin": "llm",
478
+ "status": "approved",
479
+ "superseded": false
480
+ },
481
+ {
482
+ "id": "b3",
483
+ "kind": "added",
484
+ "behavior_text": "the run check takes --continue-from for one export only and requires the directory to hold a walk; the same export under different spellings is one run; a settings file of the wrong shape is a user error naming the key",
485
+ "caused_by": [
486
+ "c13"
487
+ ],
488
+ "evidence_refs": [],
489
+ "origin": "llm",
490
+ "status": "approved",
491
+ "superseded": false
492
+ },
493
+ {
494
+ "id": "b4",
495
+ "kind": "amended",
496
+ "behavior_text": "the printed program name is evidence, the installed command; evidence explain evidence-cli still resolves to the root entry",
497
+ "caused_by": [
498
+ "c17"
499
+ ],
500
+ "evidence_refs": [],
501
+ "origin": "llm",
502
+ "status": "approved",
503
+ "superseded": false
504
+ },
505
+ {
506
+ "id": "b5",
507
+ "kind": "added",
508
+ "behavior_text": "one live test recorded in docs/notes/2026-11-18-live-test.md: six associate runs on worker and cortex, three handed back, the run check's fact counts equal to associate check's on those three",
509
+ "caused_by": [
510
+ "c11"
511
+ ],
512
+ "evidence_refs": [],
513
+ "origin": "llm",
514
+ "status": "approved",
515
+ "superseded": false
516
+ }
517
+ ],
518
+ "supersessions": []
519
+ }