okstra 0.234.1 → 0.236.0

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 (171) hide show
  1. package/dist/cli-registry.mjs +7 -0
  2. package/dist/cli-registry.mjs.map +1 -1
  3. package/dist/commands/implementation-report/implementation-report.d.mts +1 -0
  4. package/dist/commands/implementation-report/implementation-report.mjs +73 -0
  5. package/dist/commands/implementation-report/implementation-report.mjs.map +1 -0
  6. package/dist/commands/trace/comparison.d.mts +39 -0
  7. package/dist/commands/trace/comparison.mjs +138 -0
  8. package/dist/commands/trace/comparison.mjs.map +1 -0
  9. package/dist/commands/trace/document.d.mts +8 -0
  10. package/dist/commands/trace/document.mjs +37 -0
  11. package/dist/commands/trace/document.mjs.map +1 -0
  12. package/dist/commands/trace/line-counts.d.mts +15 -0
  13. package/dist/commands/trace/line-counts.mjs +107 -0
  14. package/dist/commands/trace/line-counts.mjs.map +1 -0
  15. package/dist/commands/trace/model.d.mts +2 -1
  16. package/dist/commands/trace/model.mjs +13 -1
  17. package/dist/commands/trace/model.mjs.map +1 -1
  18. package/dist/commands/trace/progression.mjs +7 -3
  19. package/dist/commands/trace/progression.mjs.map +1 -1
  20. package/dist/commands/trace/task-link.d.mts +3 -0
  21. package/dist/commands/trace/task-link.mjs.map +1 -1
  22. package/dist/commands/trace/trace.mjs +64 -44
  23. package/dist/commands/trace/trace.mjs.map +1 -1
  24. package/dist/lib/host-registry-client.d.mts +9 -2
  25. package/dist/lib/host-registry-client.mjs +29 -3
  26. package/dist/lib/host-registry-client.mjs.map +1 -1
  27. package/dist/lib/python-helper.d.mts +1 -1
  28. package/dist/lib/python-helper.mjs +11 -5
  29. package/dist/lib/python-helper.mjs.map +1 -1
  30. package/dist/lib/runtime-payload.mjs +1 -0
  31. package/dist/lib/runtime-payload.mjs.map +1 -1
  32. package/dist/lib/types.d.mts +1 -0
  33. package/docs/architecture/storage-model.md +50 -0
  34. package/docs/architecture.md +1 -1
  35. package/docs/cli.md +37 -15
  36. package/docs/project-structure-overview.md +6 -1
  37. package/package.json +1 -1
  38. package/runtime/BUILD.json +2 -2
  39. package/runtime/agents/operations/trace-tests-review.json +6 -0
  40. package/runtime/bin/lib/okstra/usage.sh +8 -8
  41. package/runtime/bin/okstra-qwen-exec.sh +5 -0
  42. package/runtime/prompts/duties/test-reviewer.json +32 -0
  43. package/runtime/prompts/lead/convergence.md +4 -1
  44. package/runtime/prompts/lead/okstra-lead-contract.md +3 -3
  45. package/runtime/prompts/lead/report-writer.md +13 -1
  46. package/runtime/prompts/profiles/_common-contract.md +10 -1
  47. package/runtime/prompts/wizard/prompts.ko.json +16 -4
  48. package/runtime/python/okstra_ctl/adapters/providers/antigravity/adapter.py +10 -2
  49. package/runtime/python/okstra_ctl/adapters/providers/claude/adapter.py +16 -1
  50. package/runtime/python/okstra_ctl/adapters/providers/codex/adapter.py +20 -5
  51. package/runtime/python/okstra_ctl/adapters/providers/kimi/adapter.py +17 -9
  52. package/runtime/python/okstra_ctl/adapters/providers/qwen/adapter.py +88 -0
  53. package/runtime/python/okstra_ctl/adapters/providers/qwen/manifest.json +1 -0
  54. package/runtime/python/okstra_ctl/adapters/providers/qwen/settings.json +12 -0
  55. package/runtime/python/okstra_ctl/adapters/providers/zai/adapter.py +8 -2
  56. package/runtime/python/okstra_ctl/agent/invocation.py +109 -56
  57. package/runtime/python/okstra_ctl/agent/prompt_cli/materialize.py +17 -0
  58. package/runtime/python/okstra_ctl/attempt_evidence.py +5 -2
  59. package/runtime/python/okstra_ctl/business_flow/contracts.py +11 -0
  60. package/runtime/python/okstra_ctl/business_flow/engine.py +90 -32
  61. package/runtime/python/okstra_ctl/business_flow/hooks.py +4 -2
  62. package/runtime/python/okstra_ctl/business_flow/source.py +39 -7
  63. package/runtime/python/okstra_ctl/business_flow/store.py +94 -28
  64. package/runtime/python/okstra_ctl/code_review_census.py +38 -0
  65. package/runtime/python/okstra_ctl/code_review_coverage.py +15 -4
  66. package/runtime/python/okstra_ctl/code_review_prepare.py +15 -3
  67. package/runtime/python/okstra_ctl/code_review_scope.py +3 -2
  68. package/runtime/python/okstra_ctl/code_review_target.py +16 -6
  69. package/runtime/python/okstra_ctl/conformance_evidence.py +44 -21
  70. package/runtime/python/okstra_ctl/consumers.py +6 -1
  71. package/runtime/python/okstra_ctl/convergence.py +95 -5
  72. package/runtime/python/okstra_ctl/convergence_critic_verify_prompt.py +11 -10
  73. package/runtime/python/okstra_ctl/convergence_engine.py +30 -15
  74. package/runtime/python/okstra_ctl/convergence_reverify_prompt.py +9 -2
  75. package/runtime/python/okstra_ctl/coverage_census.py +75 -12
  76. package/runtime/python/okstra_ctl/design_review.py +6 -26
  77. package/runtime/python/okstra_ctl/design_review_evidence.py +6 -26
  78. package/runtime/python/okstra_ctl/dispatch_core.py +9 -2
  79. package/runtime/python/okstra_ctl/dispatch_state.py +5 -0
  80. package/runtime/python/okstra_ctl/domain/role.py +1 -0
  81. package/runtime/python/okstra_ctl/entrypoints/hosts.py +1 -1
  82. package/runtime/python/okstra_ctl/execution_manifest.py +8 -0
  83. package/runtime/python/okstra_ctl/implementation_comparison/__init__.py +12 -0
  84. package/runtime/python/okstra_ctl/implementation_comparison/__main__.py +129 -0
  85. package/runtime/python/okstra_ctl/implementation_comparison/assembly.py +340 -0
  86. package/runtime/python/okstra_ctl/implementation_comparison/baseline.py +254 -0
  87. package/runtime/python/okstra_ctl/implementation_comparison/business_reuse.py +118 -0
  88. package/runtime/python/okstra_ctl/implementation_comparison/commit_source.py +113 -0
  89. package/runtime/python/okstra_ctl/implementation_comparison/html.py +49 -0
  90. package/runtime/python/okstra_ctl/implementation_comparison/inputs.py +348 -0
  91. package/runtime/python/okstra_ctl/implementation_comparison/model.py +47 -0
  92. package/runtime/python/okstra_ctl/implementation_comparison/preparation.py +195 -0
  93. package/runtime/python/okstra_ctl/implementation_comparison/publication.py +317 -0
  94. package/runtime/python/okstra_ctl/implementation_comparison/run_evidence.py +208 -0
  95. package/runtime/python/okstra_ctl/implementation_comparison/run_integration.py +198 -0
  96. package/runtime/python/okstra_ctl/implementation_comparison/source.py +342 -0
  97. package/runtime/python/okstra_ctl/implementation_comparison/target.py +144 -0
  98. package/runtime/python/okstra_ctl/implementation_comparison/task_assembly.py +333 -0
  99. package/runtime/python/okstra_ctl/implementation_comparison/task_selection.py +309 -0
  100. package/runtime/python/okstra_ctl/implementation_generation_evidence.py +134 -0
  101. package/runtime/python/okstra_ctl/implementation_outcome.py +2 -2
  102. package/runtime/python/okstra_ctl/initial_prompt_materialization.py +1 -0
  103. package/runtime/python/okstra_ctl/json_boundary.py +28 -8
  104. package/runtime/python/okstra_ctl/lead_progress.py +1 -1
  105. package/runtime/python/okstra_ctl/model_discovery.py +1 -1
  106. package/runtime/python/okstra_ctl/models.py +2 -2
  107. package/runtime/python/okstra_ctl/paths.py +7 -4
  108. package/runtime/python/okstra_ctl/phases/discussion/profile.md +2 -2
  109. package/runtime/python/okstra_ctl/phases/error_analysis/profile.md +1 -1
  110. package/runtime/python/okstra_ctl/phases/implementation_option_selection/profile.md +1 -1
  111. package/runtime/python/okstra_ctl/phases/implementation_planning/authoring.py +2 -1
  112. package/runtime/python/okstra_ctl/phases/implementation_planning/instructions/plan-body-verification.md +1 -1
  113. package/runtime/python/okstra_ctl/phases/implementation_planning/profile.md +3 -3
  114. package/runtime/python/okstra_ctl/phases/requirements_discovery/profile.md +5 -5
  115. package/runtime/python/okstra_ctl/plan_run_root.py +8 -4
  116. package/runtime/python/okstra_ctl/qa_procedures.py +33 -2
  117. package/runtime/python/okstra_ctl/registry/provider_registry.py +1 -1
  118. package/runtime/python/okstra_ctl/render.py +4 -0
  119. package/runtime/python/okstra_ctl/render_final_report.py +1 -1
  120. package/runtime/python/okstra_ctl/report_execution_evidence.py +88 -15
  121. package/runtime/python/okstra_ctl/report_finalize.py +18 -4
  122. package/runtime/python/okstra_ctl/report_html/render.py +46 -7
  123. package/runtime/python/okstra_ctl/report_html/validation.py +32 -0
  124. package/runtime/python/okstra_ctl/report_synthesis_packet.py +37 -3
  125. package/runtime/python/okstra_ctl/report_view_artifacts.py +12 -0
  126. package/runtime/python/okstra_ctl/review_verdicts.py +9 -0
  127. package/runtime/python/okstra_ctl/run.py +17 -1
  128. package/runtime/python/okstra_ctl/run_context.py +18 -3
  129. package/runtime/python/okstra_ctl/schema_excerpt.py +14 -9
  130. package/runtime/python/okstra_ctl/stage_done.py +11 -3
  131. package/runtime/python/okstra_ctl/stage_ledger.py +2 -2
  132. package/runtime/python/okstra_ctl/stage_map.py +3 -3
  133. package/runtime/python/okstra_ctl/trace_review.py +1 -0
  134. package/runtime/python/okstra_ctl/trace_tasks.py +22 -7
  135. package/runtime/python/okstra_ctl/verdict_blocks.py +7 -2
  136. package/runtime/python/okstra_ctl/verification_dispatch.py +32 -0
  137. package/runtime/python/okstra_ctl/verification_evidence.py +23 -3
  138. package/runtime/python/okstra_ctl/verification_generation.py +3 -2
  139. package/runtime/python/okstra_ctl/verification_inputs.py +3 -2
  140. package/runtime/python/okstra_ctl/verification_invocation.py +7 -3
  141. package/runtime/python/okstra_ctl/wizard/engine.py +28 -1
  142. package/runtime/python/okstra_ctl/wizard/ids.py +8 -0
  143. package/runtime/python/okstra_ctl/wizard/picker_navigation.py +64 -2
  144. package/runtime/python/okstra_ctl/wizard/roles.py +194 -45
  145. package/runtime/python/okstra_ctl/wizard/sources.py +10 -1
  146. package/runtime/python/okstra_ctl/wizard/state.py +67 -20
  147. package/runtime/python/okstra_ctl/wizard/steps_roles.py +6 -3
  148. package/runtime/python/okstra_ctl/wizard/worker_settings.py +129 -25
  149. package/runtime/python/okstra_ctl/worker_audit_ledger.py +44 -8
  150. package/runtime/python/okstra_ctl/worker_prompt_body.py +26 -2
  151. package/runtime/python/okstra_ctl/worker_prompt_headers.py +12 -9
  152. package/runtime/python/okstra_token_usage/claude.py +5 -1
  153. package/runtime/python/okstra_token_usage/collect.py +3 -3
  154. package/runtime/python/okstra_token_usage/pricing.py +1 -0
  155. package/runtime/schemas/report-synthesis-packet-v1.0.schema.json +17 -0
  156. package/runtime/skills/_fragments/wizard-relay-protocol.md +1 -1
  157. package/runtime/skills/okstra-code-review/SKILL.md +1 -1
  158. package/runtime/skills/okstra-code-review/references/census-rules.md +1 -1
  159. package/runtime/skills/okstra-code-review/references/review-calibration.md +1 -1
  160. package/runtime/skills/okstra-discussion/SKILL.md +1 -1
  161. package/runtime/skills/okstra-run/SKILL.md +33 -2
  162. package/runtime/skills/okstra-tdd/SKILL.md +74 -13
  163. package/runtime/templates/reports/html/assets/implementation-comparison.css +39 -0
  164. package/runtime/templates/reports/html/assets/implementation-comparison.js +55 -0
  165. package/runtime/templates/reports/html/i18n/en.json +113 -0
  166. package/runtime/templates/reports/html/i18n/ko.json +113 -0
  167. package/runtime/templates/reports/html/implementation-comparison.template.html +77 -0
  168. package/runtime/templates/reports/implementation-comparison-notes.md +72 -0
  169. package/runtime/templates/trace/trace.template.md +2 -0
  170. package/runtime/templates/worker-prompt-preamble.md +12 -2
  171. package/runtime/validators/validate-report-views.py +4 -36
@@ -62,7 +62,7 @@ def build_stage_ledger(task_root: Path) -> dict[str, Any] | None:
62
62
  records = lifecycle.ledger_records()
63
63
 
64
64
  built_from, divergence = _built_from(task_root, latest)
65
- divergence += _renumbering_divergence(
65
+ divergence += renumbering_divergence(
66
66
  latest, built_from, state.done_stages
67
67
  )
68
68
  ledger: dict[str, Any] = {
@@ -112,7 +112,7 @@ def _built_from(task_root: Path, latest: Any):
112
112
  ]
113
113
 
114
114
 
115
- def _renumbering_divergence(
115
+ def renumbering_divergence(
116
116
  latest: Any,
117
117
  built_from: Any,
118
118
  done_stages: set[int],
@@ -378,8 +378,8 @@ def stage_row_is_cancelled(
378
378
  return status == "cancelled"
379
379
 
380
380
 
381
- def _parse_schema_v2_stage_map(
382
- data: dict[str, Any], source_plan_path: str,
381
+ def parse_stage_map_data(
382
+ data: dict[str, Any], *, source_plan_path: str = "",
383
383
  ) -> list[StageMapStage]:
384
384
  planning = data.get("implementationPlanning")
385
385
  stage_map = planning.get("stageMap") if isinstance(planning, dict) else None
@@ -416,7 +416,7 @@ def parse_stage_map_file(plan_path: Path) -> list[StageMapStage]:
416
416
  raise StageMapError(
417
417
  "stage_map", "structured report must be an object", str(resolved)
418
418
  )
419
- return _parse_schema_v2_stage_map(data, str(resolved))
419
+ return parse_stage_map_data(data, source_plan_path=str(resolved))
420
420
  return _parse_stage_map_markdown(resolved)
421
421
 
422
422
 
@@ -34,6 +34,7 @@ STEP_OPERATIONS = {
34
34
  "analysis-review": "trace-scope-review",
35
35
  "spec-review": "trace-scope-review",
36
36
  "interface-review": "trace-interface-review",
37
+ "tests-review": "trace-tests-review",
37
38
  }
38
39
  _DIGEST_RE = re.compile(r"^sha256:[0-9a-f]{64}$")
39
40
  _VERDICT_RE = re.compile(r"^Trace review verdict: (pass|fail)[ \t]*$", re.MULTILINE)
@@ -6,6 +6,7 @@ import hashlib
6
6
  import json
7
7
  import math
8
8
  import re
9
+ from collections.abc import Mapping
9
10
  from dataclasses import dataclass
10
11
  from datetime import datetime, timezone
11
12
  from pathlib import Path
@@ -30,7 +31,7 @@ class TraceSource:
30
31
  digest: str
31
32
 
32
33
 
33
- def read_task_trace(project_root: Path, trace_path: str, task_key: str, *, budget: ReadBudget | None = None) -> TraceSource:
34
+ def read_trace_source(project_root: Path, trace_path: str, *, budget: ReadBudget | None = None) -> TraceSource:
34
35
  root = project_root.resolve()
35
36
  relative = Path(trace_path)
36
37
  if relative.is_absolute() or len(relative.parts) != 4 or relative.parts[:2] != (*OKSTRA_RELATIVE.parts, "traces"):
@@ -38,14 +39,31 @@ def read_task_trace(project_root: Path, trace_path: str, task_key: str, *, budge
38
39
  if trace_path != relative.as_posix() or contained_relative_path(trace_path, root) != trace_path or relative.suffix != ".md":
39
40
  raise ValueError("Trace path is not canonical or contained")
40
41
  content = read_bounded_bytes(root / relative, max_bytes=TASK_MANIFEST_MAX_BYTES, artifact="implementation trace", budget=budget)
42
+ return parse_trace_source(content)
43
+
44
+
45
+ def parse_trace_source(content: bytes) -> TraceSource:
41
46
  text = content.decode("utf-8")
42
47
  match = re.match(r"^---\r?\n([\s\S]*?)\r?\n---(?:\r?\n|$)", text)
43
48
  header = json.loads(match[1]) if match else None
44
- if not isinstance(header, dict) or not isinstance(header.get("task"), dict) or header["task"].get("key") != task_key:
45
- raise ValueError("Trace owner does not match the task")
49
+ if not isinstance(header, dict):
50
+ raise ValueError("Trace frontmatter must be a JSON object")
46
51
  return TraceSource(header, text, hashlib.sha256(content).hexdigest())
47
52
 
48
53
 
54
+ def read_task_trace(project_root: Path, trace_path: str, task_key: str, *, budget: ReadBudget | None = None) -> TraceSource:
55
+ source = read_trace_source(project_root, trace_path, budget=budget)
56
+ if not isinstance(source.frontmatter.get("task"), dict) or source.frontmatter["task"].get("key") != task_key:
57
+ raise ValueError("Trace owner does not match the task")
58
+ return source
59
+
60
+
61
+ def trace_integration_completed(frontmatter: Mapping[str, object]) -> bool:
62
+ steps = frontmatter.get("steps", {})
63
+ integration = steps.get("integration", {}) if isinstance(steps, dict) else {}
64
+ return isinstance(integration, dict) and integration.get("status") == "done"
65
+
66
+
49
67
  def _validate_request(project_root: Path, request: dict) -> dict:
50
68
  if not isinstance(request, dict) or len(json.dumps(request, ensure_ascii=False).encode()) > TRACE_REQUEST_MAX_BYTES:
51
69
  raise ValueError("Trace request exceeds its input limit")
@@ -71,10 +89,7 @@ def _validate_request(project_root: Path, request: dict) -> dict:
71
89
  raise ValueError("Trace creation timestamp does not match its source")
72
90
  if source.digest != request.get("digest") or source.frontmatter.get("currentStep") != request["currentStep"]:
73
91
  raise ValueError("Trace projection does not match its saved source")
74
- steps = source.frontmatter.get("steps", {})
75
- integration = steps.get("integration", {}) if isinstance(steps, dict) else {}
76
- completed = isinstance(integration, dict) and integration.get("status") == "done"
77
- if request["completed"] != completed:
92
+ if request["completed"] != trace_integration_completed(source.frontmatter):
78
93
  raise ValueError("Trace completion projection does not match its source")
79
94
  task = source.frontmatter["task"]
80
95
  if set(task) != {"key", *TRACE_METADATA_LIST_LIMITS, *TRACE_METADATA_TEXT_LIMITS}:
@@ -65,7 +65,7 @@ _ITEM_RE = re.compile(r"^###[ \t]+(?P<id>[^\s:]+)[ \t]*:?.*$", re.MULTILINE)
65
65
  # `_field_match` enforces that an opening `**` is closed exactly once and that
66
66
  # a label without bold carries a bullet.
67
67
  _FIELD_KEYS = ("Verdict", "Fixability", "Note", "Explanation", "Prior dissent",
68
- "Basis", "Your evidence", "Dependencies")
68
+ "Basis", "Your evidence", "Dependencies", "Opened")
69
69
  _CANONICAL_KEYS = {key.lower(): key for key in _FIELD_KEYS}
70
70
  _FIELD_RE = re.compile(
71
71
  r"^(?P<bullet>[-*+][ \t]+)?(?P<open>\*\*)?"
@@ -105,6 +105,7 @@ class FindingVote:
105
105
  verdict: str
106
106
  disagree_basis: str | None
107
107
  explanation: str
108
+ opened: tuple[str, ...] = ()
108
109
 
109
110
 
110
111
  def _scan_blocks(text: str) -> dict[str, dict[str, str]]:
@@ -221,7 +222,11 @@ def _finding_vote(
221
222
  f"finding `{finding_id}` has basis `{basis}` — expected one of "
222
223
  f"{sorted(DISAGREE_BASES)}"
223
224
  )
224
- return FindingVote(finding_id, verdict, basis, explanation)
225
+ from .worker_audit_ledger import named_files
226
+
227
+ return FindingVote(
228
+ finding_id, verdict, basis, explanation, named_files(fields.get("Opened", "")),
229
+ )
225
230
 
226
231
 
227
232
  @dataclass(frozen=True)
@@ -6,6 +6,9 @@ from typing import Any, Mapping
6
6
 
7
7
  from .correction_round import completed_round_attempts, correction_round
8
8
  from .execution_manifest import read_execution_manifest
9
+ from .execution_identity import ExecutionManifest, build_dispatch_id
10
+ from .attempt_evidence import AttemptEvidenceSeal, read_sealed_artifact
11
+ from .json_boundary import ArtifactByteReader
9
12
  from .paths import RunRef
10
13
  from .qa_procedures import safe_qa_path
11
14
  from .verification_generation import prepare_verification_generation
@@ -60,6 +63,35 @@ def generation_sealed_after_round(round_number: int) -> int:
60
63
  return 1 + round_number
61
64
 
62
65
 
66
+ def implementation_generation_frontier(execution: ExecutionManifest) -> int | None:
67
+ return max((verification_generation_number(row.dispatch_kind)
68
+ for row in execution.invocations
69
+ if row.duty_id in ("implementation-executor", "implementation-verifier")), default=None)
70
+
71
+
72
+ def read_completed_attempt_result(
73
+ project: Path, execution: ExecutionManifest, invocation_ref: str, attempt: int,
74
+ *, reader: ArtifactByteReader | None = None,
75
+ ) -> tuple[Path, bytes]:
76
+ row = next((row for row in execution.attempts if (row.invocation_ref, row.attempt)
77
+ == (invocation_ref, attempt)), None)
78
+ if row is None or row.status != "ok" or row.finished_at is None or not row.result_path:
79
+ raise ValueError("result requires the current completed ok attempt")
80
+ attempt_ref = build_dispatch_id(invocation_ref, attempt)
81
+ seals = [seal for seal in execution.attempt_evidence_seals
82
+ if seal.get("sealRef") == row.evidence_seal_ref and seal.get("attemptRef") == attempt_ref]
83
+ if len(seals) != 1:
84
+ raise ValueError("result requires its attempt evidence seal")
85
+ path = Path(row.result_path)
86
+ relative = path.relative_to(project).as_posix() if path.is_absolute() else path.as_posix()
87
+ if not relative.startswith(".okstra/"):
88
+ raise ValueError("result must remain in project artifacts")
89
+ path = safe_qa_path(project, relative)
90
+ seal = AttemptEvidenceSeal.from_payload(seals[0])
91
+ raw = read_sealed_artifact(seal, path) if reader is None else read_sealed_artifact(seal, path, reader=reader)
92
+ return path, raw
93
+
94
+
63
95
  def _seal(
64
96
  project: Path, manifest_path: Path, manifest: Mapping[str, Any],
65
97
  active_context: Mapping[str, Any], *, audience: str, dispatch_kind: str,
@@ -23,6 +23,7 @@ from .execution_mutation_audit import (
23
23
  compare_verification_sources,
24
24
  )
25
25
  from .json_boundary import (
26
+ ArtifactByteReader, OwnedJsonSnapshot, ReadBudget,
26
27
  JsonBoundaryError, load_owned_object, load_owned_object_snapshot,
27
28
  write_owned_object_atomic,
28
29
  )
@@ -308,6 +309,17 @@ def roster_verifier_role_execution_ref(
308
309
  return None
309
310
 
310
311
 
312
+ def implementation_verifier_roster(manifest: Mapping[str, Any]) -> list[str]:
313
+ assignments = manifest.get("workerAssignments", [])
314
+ if not isinstance(assignments, list) or any(not isinstance(row, Mapping) for row in assignments):
315
+ raise ValueError("implementation verifier roster is malformed")
316
+ roster = [row.get("workerId") for row in assignments
317
+ if row.get("role") == "verifier"]
318
+ if not roster or any(not isinstance(worker, str) or not worker for worker in roster) or len(set(roster)) != len(roster):
319
+ raise ValueError("implementation requires a distinct verifier roster")
320
+ return roster
321
+
322
+
311
323
  def _sealed_conformance_entry(
312
324
  project: Path, ref: RunRef, generation: VerificationGenerationRecord,
313
325
  check_ref: str, stage_key: str,
@@ -647,10 +659,18 @@ def _publish(check_run_dir: Path, record: VerificationExecutionRecord) -> None:
647
659
 
648
660
 
649
661
  def read_verification_execution_record(path: Path) -> VerificationExecutionRecord:
650
- return cast(VerificationExecutionRecord, load_owned_object(
662
+ return cast(VerificationExecutionRecord, read_verification_execution_snapshot(path).value)
663
+
664
+
665
+ def read_verification_execution_snapshot(
666
+ path: Path, *, reader: ArtifactByteReader | None = None,
667
+ max_bytes: int | None = None, budget: ReadBudget | None = None,
668
+ ) -> OwnedJsonSnapshot:
669
+ return load_owned_object_snapshot(
651
670
  path, artifact="verification execution record", schema=_RECORD_SCHEMA,
652
- validate_cross_fields=_record_errors,
653
- ))
671
+ validate_cross_fields=_record_errors, reader=reader,
672
+ max_bytes=max_bytes, budget=budget,
673
+ )
654
674
 
655
675
 
656
676
  def _record_errors(record: dict[str, Any]) -> list[str]:
@@ -12,7 +12,7 @@ from .execution_mutation_audit import (
12
12
  VerificationSourceSnapshot, capture_verification_source,
13
13
  is_generated_verification_path,
14
14
  )
15
- from .json_boundary import load_owned_object_snapshot
15
+ from .json_boundary import ArtifactByteReader, load_owned_object_snapshot
16
16
  from .paths import RunRef
17
17
  from .qa_commands import (
18
18
  QA_COMMAND_EXECUTING_TASK_TYPES, QA_SOURCE_DIGEST_PATTERN, qa_relative_path_valid,
@@ -267,11 +267,12 @@ def sealed_source_snapshot(
267
267
 
268
268
 
269
269
  def read_verification_generation(
270
- path: Path, expected_digest: str,
270
+ path: Path, expected_digest: str, *, reader: ArtifactByteReader | None = None,
271
271
  ) -> VerificationGenerationRecord:
272
272
  stored = load_owned_object_snapshot(
273
273
  path, artifact=_ARTIFACT, schema=_SCHEMA,
274
274
  validate_cross_fields=_generation_errors,
275
+ reader=reader,
275
276
  )
276
277
  if content_digest(stored.raw_bytes) != expected_digest:
277
278
  raise ValueError("verification generation digest mismatch")
@@ -8,7 +8,7 @@ from typing import Any, Callable, Mapping, cast
8
8
 
9
9
  from okstra_project.dirs import OKSTRA_DIR_NAME, json_lock_path, registry_lock
10
10
 
11
- from .json_boundary import load_owned_object_snapshot, write_owned_object_atomic
11
+ from .json_boundary import ArtifactByteReader, load_owned_object_snapshot, write_owned_object_atomic
12
12
  from .paths import RunRef
13
13
  from .review_policy import capture_review_policy, review_policy_errors
14
14
  from .qa_commands import (
@@ -218,10 +218,11 @@ def capture_verification_inputs(
218
218
  return payload
219
219
 
220
220
 
221
- def read_verification_inputs(path: Path, expected_digest: str) -> VerificationInputs:
221
+ def read_verification_inputs(path: Path, expected_digest: str, *, reader: ArtifactByteReader | None = None) -> VerificationInputs:
222
222
  snapshot = load_owned_object_snapshot(
223
223
  path, artifact=INPUT_ARTIFACT, schema=INPUT_SCHEMA,
224
224
  validate_cross_fields=_verification_input_errors,
225
+ reader=reader,
225
226
  )
226
227
  if content_digest(snapshot.raw_bytes) != expected_digest:
227
228
  raise ValueError("verification input digest mismatch")
@@ -17,6 +17,7 @@ from .verification_generation import (
17
17
  from .verification_input_ref import VerificationInputReference
18
18
  from .verification_inputs import read_verification_inputs
19
19
  from .verification_input_types import DESIGN_REVIEW_INPUT_SCHEMA_VERSION
20
+ from .json_boundary import ArtifactByteReader
20
21
 
21
22
  VERIFICATION_EVIDENCE_VERSION = "1.0"
22
23
  VERIFICATION_GENERATION_HEADER_PREFIXES = (
@@ -63,13 +64,14 @@ def _has_c0_control_characters(value: str) -> bool:
63
64
 
64
65
  def read_invocation_generation(
65
66
  project: Path, manifest_path: Path | None, manifest: Mapping[str, Any],
66
- reference: VerificationGenerationReference | None, audience: str,
67
+ reference: VerificationGenerationReference | None, audience: str, *,
68
+ reader: ArtifactByteReader | None = None,
67
69
  ) -> VerificationGenerationRecord | None:
68
70
  if not _generation_required(manifest, reference, audience):
69
71
  return None
70
72
  if reference is None or manifest_path is None:
71
73
  raise ValueError("verifier invocation requires a verification generation")
72
- return _read_bound_generation(project, manifest_path, manifest, reference)
74
+ return _read_bound_generation(project, manifest_path, manifest, reference, reader=reader)
73
75
 
74
76
 
75
77
  def verification_generation_required(
@@ -102,11 +104,12 @@ def _generation_required(
102
104
  def _read_bound_generation(
103
105
  project: Path, manifest_path: Path, manifest: Mapping[str, Any],
104
106
  reference: VerificationGenerationReference,
107
+ *, reader: ArtifactByteReader | None = None,
105
108
  ) -> VerificationGenerationRecord:
106
109
  task_type = manifest["taskType"]
107
110
  project = project.resolve()
108
111
  path = safe_qa_path(project, reference.path)
109
- record = read_verification_generation(path, reference.digest)
112
+ record = read_verification_generation(path, reference.digest, **({"reader": reader} if reader is not None else {}))
110
113
  relative_manifest = manifest_path.resolve().relative_to(project).as_posix()
111
114
  if (record["runManifestPath"] != relative_manifest
112
115
  or record["taskType"] != task_type):
@@ -116,6 +119,7 @@ def _read_bound_generation(
116
119
  raise ValueError("verification generation input reference differs from run")
117
120
  inputs = read_verification_inputs(
118
121
  safe_qa_path(project, expected.relative_path), expected.digest,
122
+ **({"reader": reader} if reader is not None else {}),
119
123
  )
120
124
  if record["sourceRoot"] != inputs["sourceRoot"]:
121
125
  raise ValueError("verification generation source differs from frozen inputs")
@@ -37,11 +37,12 @@ from .ids import (
37
37
  _STEP_TO_GROUP,
38
38
  )
39
39
  from .state import Prompt, WizardError, WizardState, _is_role_selection_step
40
- from .prompts import _domain_prompt
40
+ from .prompts import _domain_prompt, _msg
41
41
  from .picker_navigation import (
42
42
  is_split_picker,
43
43
  merge_split_picker_answer,
44
44
  present_picker,
45
+ split_by_vendor,
45
46
  split_picker,
46
47
  )
47
48
  from .roles import _step_role, _submit_role_prompt, next_role_prompt
@@ -184,9 +185,19 @@ def next_prompt(state: WizardState) -> Prompt:
184
185
  state.__dict__.update(attempt.__dict__)
185
186
 
186
187
 
188
+ def _is_model_pick(prompt: Prompt) -> bool:
189
+ return prompt.kind == "pick" and (
190
+ prompt.step.startswith(("role-models:", "role-model:"))
191
+ or prompt.step == "worker_settings_models"
192
+ )
193
+
194
+
187
195
  def _native_picker_screen(state: WizardState, prompt: Prompt) -> Prompt:
188
196
  """호스트 네이티브 선택기 한도에 맞춘 화면.
189
197
 
198
+ 모델 선택 화면은 후보 수와 상관없이 벤더별 탭으로 나눈다(`split_by_vendor`).
199
+ 그 묶음이 호스트에 실리지 않을 때만 아래 규칙을 쓴다.
200
+
190
201
  한도를 넘는 픽은 체크박스든 단일 선택이든 네이티브 질문 묶음에 실리는
191
202
  크기(claude-code 4×4=16)까지 같은 화면의 체크박스 질문 여러 개로 자른다
192
203
  (`split_picker`) — 사용자는 탭을 옮겨 다니며 한 번에 답한다. 단일 선택을
@@ -204,6 +215,16 @@ def _native_picker_screen(state: WizardState, prompt: Prompt) -> Prompt:
204
215
  return prompt
205
216
  prompt = prompt.questions[0]
206
217
  port = default_host_registry().resolve(state.host_runtime).interaction()
218
+ if _is_model_pick(prompt):
219
+ by_vendor = split_by_vendor(
220
+ prompt,
221
+ max_options=port.native_option_limit,
222
+ max_questions=port.native_question_limit,
223
+ free_input_label=_msg(state.workspace_root, "model_tabs", "free_input", vendors="{vendors}"),
224
+ mixed_label=_msg(state.workspace_root, "model_tabs", "mixed", vendors="{vendors}"),
225
+ )
226
+ if by_vendor is not prompt and _interaction_plan(state, by_vendor).kind == "native-group":
227
+ return by_vendor
207
228
  split = split_picker(
208
229
  prompt,
209
230
  max_options=port.native_option_limit,
@@ -276,6 +297,12 @@ def _sim_advance(state: WizardState, prompt: Prompt) -> None:
276
297
  prompt = _next_prompt_screen(state)
277
298
  try:
278
299
  if _is_role_selection_step(prompt.step):
300
+ # A pending typed-model step has no default answer: count it and
301
+ # settle the role from its model screen's recommended set.
302
+ if prompt.kind == "text":
303
+ state.role_model_drafts.pop(_step_role(prompt.step), None)
304
+ state.answered.append(prompt.step)
305
+ prompt = next_role_prompt(state) or prompt
279
306
  # 화면은 호스트 한도에 맞춰 쪽으로 잘린 사본일 수 있다 — 기본답은
280
307
  # 잘리지 않은 원본(추천 집합)에서 낸다.
281
308
  original = next_role_prompt(state) or prompt
@@ -348,8 +348,16 @@ _REUSE_FILLED_STEPS = (
348
348
  S_CLARIFICATION_PICK, S_CLARIFICATION)
349
349
 
350
350
 
351
+ _INPUT_PREFIX = "role-model-input:"
351
352
  _MODELS_PREFIX = "role-models:"
352
353
  _SINGLE_PREFIX = "role-model:"
354
+ _ROLE_SELECTION_PREFIXES = (_INPUT_PREFIX, _MODELS_PREFIX, _SINGLE_PREFIX)
355
+ # A vendor tab's "type a model" row: the prefix plus the tab's providers joined by "+".
356
+ MODEL_FREE_INPUT_PREFIX = "__free_input__:"
357
+
358
+
359
+ def _role_model_input_prompt_id(role: str) -> str:
360
+ return f"{_INPUT_PREFIX}{role}"
353
361
 
354
362
 
355
363
  def _role_models_prompt_id(role: str) -> str:
@@ -13,8 +13,8 @@ import json
13
13
  import math
14
14
  from dataclasses import replace
15
15
 
16
- from .ids import PICK_TYPE_CUSTOM
17
- from .state import Prompt, WizardError
16
+ from .ids import MODEL_FREE_INPUT_PREFIX, PICK_TYPE_CUSTOM
17
+ from .state import Option, Prompt, WizardError
18
18
 
19
19
  _SPLIT_SEPARATOR = "#"
20
20
 
@@ -69,6 +69,68 @@ def split_picker(
69
69
  )
70
70
 
71
71
 
72
+ def split_by_vendor(
73
+ prompt: Prompt,
74
+ *,
75
+ max_options: int,
76
+ max_questions: int,
77
+ free_input_label: str,
78
+ mixed_label: str,
79
+ ) -> Prompt:
80
+ """Show a model pick as one checkbox tab per vendor on one screen.
81
+
82
+ Each tab lists up to `max_options - 1` of a vendor's models, keeping the
83
+ recommended ones, and ends with a "type a model" row whose value names the tab's
84
+ providers. A vendor with that many models gets its own tab; the others
85
+ share one mixed tab. When the own tabs and the mixed tab exceed
86
+ `max_questions`, the last own tabs join the mixed one. Rows that are not
87
+ `provider/model` refs (a skip row) are left out: an empty answer means
88
+ nothing picked. Returns the prompt unchanged when it holds no model.
89
+ """
90
+ models = [option for option in prompt.options if "/" in option.value]
91
+ if prompt.kind != "pick" or not models or max_options < 2:
92
+ return prompt
93
+ shown = max_options - 1
94
+ vendors: dict[str, list[Option]] = {}
95
+ for option in models:
96
+ vendors.setdefault(option.value.split("/", 1)[0], []).append(option)
97
+ own = [vendor for vendor, rows in vendors.items() if len(rows) >= shown]
98
+ mixed = [vendor for vendor in vendors if vendor not in own]
99
+ if len(own) + (1 if mixed else 0) > max_questions:
100
+ keep = max_questions - 1
101
+ mixed = [vendor for vendor in vendors if vendor in own[keep:] or vendor in mixed]
102
+ own = own[:keep]
103
+ groups = [[vendor] for vendor in own] + ([mixed] if mixed else [])
104
+ questions: list[Prompt] = []
105
+ for index, group in enumerate(groups, start=1):
106
+ rows = [option for vendor in group for option in vendors[vendor]]
107
+ picked = sorted(rows, key=lambda option: not option.recommended)[:shown]
108
+ rows = [option for option in rows if option in picked]
109
+ name = group[0] if len(group) == 1 else mixed_label.format(vendors="·".join(group))
110
+ questions.append(replace(
111
+ prompt,
112
+ step=f"{prompt.step}{_SPLIT_SEPARATOR}{index}",
113
+ label=f"{prompt.label} [{name}]",
114
+ options=[
115
+ *rows,
116
+ Option(
117
+ f"{MODEL_FREE_INPUT_PREFIX}{'+'.join(group)}",
118
+ free_input_label.format(vendors=name),
119
+ ),
120
+ ],
121
+ multi=True,
122
+ ))
123
+ return Prompt(
124
+ step=prompt.step,
125
+ kind="pick_group",
126
+ label=prompt.label,
127
+ help=prompt.help,
128
+ echo_template=prompt.echo_template,
129
+ multi=prompt.multi,
130
+ questions=questions,
131
+ )
132
+
133
+
72
134
  def is_split_picker(prompt: Prompt) -> bool:
73
135
  return prompt.kind == "pick_group" and bool(prompt.questions) and all(
74
136
  question.multi