@tea-agent/loop-agent 0.33.6 → 0.33.7-beta.1

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 (69) hide show
  1. package/CHANGELOG.md +39 -23
  2. package/dist/application/task-lifecycle/advance.js +4 -254
  3. package/dist/application/task-lifecycle/gates.js +0 -50
  4. package/dist/application/task-lifecycle/observe.js +2 -11
  5. package/dist/commands/init-upgrade.js +1 -32
  6. package/dist/commands/init.js +3 -94
  7. package/dist/executors/shell-executor.js +91 -4
  8. package/dist/executors/shell-write-guard.js +8 -26
  9. package/dist/shared/operator/capabilities.js +42 -72
  10. package/dist/task/source-prepare/index.js +0 -2
  11. package/dist/task/source-prepare/parse-intent.js +10 -58
  12. package/dist/task/source-prepare/prepare.js +16 -180
  13. package/dist/task/source-prepare/reference-integrity.js +2 -18
  14. package/dist/worker/console/app-data.js +0 -2
  15. package/dist/worker/console/chat/chat-event-store.js +25 -190
  16. package/dist/worker/console/chat/instruction-skills.js +217 -0
  17. package/dist/worker/console/chat/pi-console-config.js +32 -250
  18. package/dist/worker/console/chat/pi-runtime.js +71 -625
  19. package/dist/worker/console/chat/resource-loader.js +4 -5
  20. package/dist/worker/console/chat/routes.js +146 -324
  21. package/dist/worker/console/chat/runtime-context.js +12 -48
  22. package/dist/worker/console/chat/runtime-selection.js +0 -59
  23. package/dist/worker/console/chat/shortcuts.js +0 -1
  24. package/dist/worker/console/chat/tool-adapter.js +3 -9
  25. package/dist/worker/console/chat/tools.js +1 -5
  26. package/dist/worker/console/operator-actions.js +68 -559
  27. package/dist/worker/console/server.js +15 -8
  28. package/dist/worker/console/static/assets/index-CnUXAqxG.css +1 -0
  29. package/dist/worker/console/static/assets/index-CteJFFL2.js +29 -0
  30. package/dist/worker/console/static/index.html +2 -2
  31. package/dist/worker/console/static-src/operator-chat/chat-sse-events.js +8 -45
  32. package/dist/worker/console/static-src/operator-chat/refs.js +0 -9
  33. package/dist/worker/console/static-src/operator-chat/useChatSessions.js +0 -16
  34. package/dist/worker/console/static-src/operator-chat/useChatStream.js +184 -210
  35. package/dist/worker/console/static-src/operator-chat/useChatThread.js +5 -49
  36. package/dist/worker/console/static-src/operator-chat/useComposer.js +0 -17
  37. package/dist/worker/console/static-src/operator-chat/useRuntimeControls.js +74 -225
  38. package/dist/worker/delivery/final-verification.js +5 -13
  39. package/dist/worker/delivery/package.js +19 -31
  40. package/dist/worker/delivery/verification-bundle.js +4 -6
  41. package/dist/worker/observe/static/operator-chrome.css +2 -5
  42. package/dist/worker/observe/static/operator-chrome.js +1 -6
  43. package/dist/worker/observe/static/styles.css +9 -39
  44. package/dist/workflows/dag/backend-test-case-coverage-analysis.js +462 -33
  45. package/dist/workflows/dag/backend-test-case-manifest.js +4 -0
  46. package/dist/workflows/dag/backend-test-markdown-workflow.js +25 -1
  47. package/dist/workflows/dag/backend-test-module-stem.js +5 -0
  48. package/dist/workflows/dag/backend-test-pytest-collection.js +345 -24
  49. package/dist/workflows/dag/backend-test-scenario-param.js +552 -124
  50. package/dist/workflows/dag/backend-test-writer-completeness.js +47 -16
  51. package/dist/workflows/dag/dynamic-runtime/map.js +24 -8
  52. package/dist/workflows/dag/frontend-worktree-diff.js +27 -12
  53. package/dist/workflows/dag/init-hybrid.js +49 -37
  54. package/dist/workflows/dag/types.js +7 -0
  55. package/dist/workflows/dag/workspace-checkpoint.js +27 -8
  56. package/docs/templates/backend-test-dag.json +33 -30
  57. package/harness.json +1 -1
  58. package/package.json +1 -1
  59. package/skills/loop-agent/references/command-reference.md +1 -3
  60. package/skills/loop-agent/references/source-and-plan-practice.md +0 -13
  61. package/skills/loop-agent/references/task-workflow.md +0 -4
  62. package/dist/shared/resilient-git.js +0 -133
  63. package/dist/task/source-prepare/artifact-meta.js +0 -137
  64. package/dist/task/source-prepare/semantic-intake.js +0 -404
  65. package/dist/worker/console/dag-execution-receipt.js +0 -380
  66. package/dist/worker/console/static/assets/index-BUOLppPr.js +0 -28
  67. package/dist/worker/console/static/assets/index-C1KzazY5.css +0 -1
  68. package/dist/worker/console/static-src/operator-chat/runtime-snapshot-store.js +0 -257
  69. package/dist/worker/console/static-src/operator-chat/useRuntimeSnapshot.js +0 -196
@@ -1,4 +1,5 @@
1
1
  import { createHash } from "node:crypto";
2
+ import { spawnSync } from "node:child_process";
2
3
  import { access, mkdir, readFile, readdir, writeFile } from "node:fs/promises";
3
4
  import path from "node:path";
4
5
  import { z } from "zod";
@@ -8,7 +9,15 @@ const CASE_HEADING = /^##\s+(BE-[A-Z0-9_-]+-\d{3})\b.*$/gm;
8
9
  const CASE_ID_IN_TEXT = /\bBE-[A-Z0-9_-]+-\d{3}\b/g;
9
10
  const NON_CANONICAL_CASE_HEADING = /^##\s+(BE-[A-Z0-9_-]+-(?:\d{2}|\d{2,3}[A-Z]+))\b.*$/gm;
10
11
  const AC_ID_IN_TEXT = /\bAC-[A-Z0-9]+(?:-[A-Z0-9]+)*\b/g;
11
- const RULE_KEY = /^(?:AC|REQ|BR|API|GET|POST|PUT|PATCH|DELETE)-[A-Z0-9]+(?:-[A-Z0-9]+)*$/;
12
+ const SOURCE_RULE_ID_IN_TEXT = /\b(?:AC|REQ|BR)-[A-Z0-9]+(?:-[A-Z0-9]+)*\b/g;
13
+ const SOURCE_ERROR_CODE_IN_TEXT = /\b[A-Z][0-9]{5,}\b/g;
14
+ const JSON_FIELD_IN_TEXT = /["']([A-Za-z][A-Za-z0-9_]*)["']\s*:/g;
15
+ const DTO_FIELD_IN_TEXT = /\b(?:private|public|protected)\s+(?:static\s+)?(?:final\s+)?[A-Za-z_$][A-Za-z0-9_$<>,.?\[\] ]*\s+([a-z][A-Za-z0-9_]*)\s*[;=]/g;
16
+ const NATURAL_BUSINESS_RULE_KEYWORDS = /(?:事务|原子性|回滚|幂等|唯一性|唯一|状态转换|状态流转|权限|角色|租户|跨字段|transaction|atomic|rollback|idempoten|unique|state transition|authorization|permission|tenant|cross-field)/i;
17
+ const NATURAL_BUSINESS_RULE_NORMATIVE = /(?:必须|不得|禁止|仅当|应当|需要|不能|保证|must|shall|required|must not|only when|guarantee)/i;
18
+ // Token form is used for section scanning; anchored form validates full cells.
19
+ const RULE_KEY_TOKEN = /(?:AC|REQ|BR|API|GET|POST|PUT|PATCH|DELETE)-[A-Z0-9]+(?:-[A-Z0-9]+)*|[A-Z][0-9]{5,}/;
20
+ const RULE_KEY = new RegExp(`^(?:${RULE_KEY_TOKEN.source})$`);
12
21
  const TEST_POINT = /^TP-[A-Z0-9]+(?:-[A-Z0-9]+)*$/;
13
22
  const COVERAGE_HEADERS = [
14
23
  "Rule Key",
@@ -87,6 +96,13 @@ const coverageCaseSchema = z.object({
87
96
  scenarioTypes: z.array(z.string()),
88
97
  declaredScripts: z.array(z.string()),
89
98
  declaredPrimarySymbols: z.array(z.string()),
99
+ payloadContract: z.object({
100
+ declared: z.boolean(),
101
+ noPayload: z.boolean(),
102
+ requiredPaths: z.array(z.string()),
103
+ allowedPaths: z.array(z.string()),
104
+ enumValues: z.record(z.string(), z.array(z.string())),
105
+ }).strict(),
90
106
  }).strict();
91
107
  const coverageRuleSchema = z.object({
92
108
  ruleKey: z.string().min(1),
@@ -102,6 +118,23 @@ const coverageRuleSchema = z.object({
102
118
  missingTestPoints: z.array(z.string()),
103
119
  status: z.enum(["COVERED", "PARTIAL", "GAP", "CONFLICT"]),
104
120
  }).strict();
121
+ const sourceCoverageFactSchema = z.object({
122
+ kind: z.enum(["requirement-id", "error-code", "dto-field", "business-rule"]),
123
+ key: z.string().min(1),
124
+ sourcePath: z.string().min(1),
125
+ sourceSha256: sha256Schema,
126
+ lineStart: z.number().int().positive(),
127
+ lineEnd: z.number().int().positive(),
128
+ quote: z.string().min(1).max(1000),
129
+ covered: z.boolean(),
130
+ }).strict();
131
+ const sourceCoverageSchema = z.object({
132
+ sourceFactCount: z.number().int().min(0),
133
+ coveredSourceFactCount: z.number().int().min(0),
134
+ residualSourceFactCount: z.number().int().min(0),
135
+ facts: z.array(sourceCoverageFactSchema),
136
+ residual: z.array(sourceCoverageFactSchema),
137
+ }).strict();
105
138
  const coverageScopeSchema = z.object({
106
139
  changeClassification: z.enum(CHANGE_CLASSIFICATIONS),
107
140
  coveragePolicy: z.enum(COVERAGE_POLICIES),
@@ -112,13 +145,14 @@ const coverageScopeSchema = z.object({
112
145
  completenessClaim: z.enum(["affected-operations-full", "affected-scope"]),
113
146
  }).strict();
114
147
  export const backendTestCaseCoverageFactsSchema = z.object({
115
- schemaId: z.literal("backend-test-case-coverage-facts-v3"),
116
- schemaVersion: z.literal(3),
148
+ schemaId: z.literal("backend-test-case-coverage-facts-v4"),
149
+ schemaVersion: z.literal(4),
117
150
  taskId: z.string().min(1),
118
151
  status: z.enum(["PASS", "FAIL", "UNAVAILABLE"]),
119
152
  sourceBinding: sourceBindingSchema,
120
153
  inputFiles: z.array(inputFileSchema),
121
154
  coverageScope: coverageScopeSchema,
155
+ sourceCoverage: sourceCoverageSchema,
122
156
  summary: z.object({
123
157
  explicitAcCount: z.number().int().min(0),
124
158
  coveredAcCount: z.number().int().min(0),
@@ -177,12 +211,24 @@ const correspondenceEntrySchema = z.object({
177
211
  "CROSS_CUTTING_EVIDENCE_MISSING",
178
212
  "TEST_POINT_BINDING_DUPLICATE",
179
213
  "TEST_POINT_BINDING_EXTRA",
214
+ "PAYLOAD_CONTRACT_UNAVAILABLE",
215
+ "PAYLOAD_SHAPE_MISMATCH",
180
216
  ]),
217
+ payloadAssessment: z.object({
218
+ status: z.enum(["SAFE", "UNSAFE", "UNAVAILABLE"]),
219
+ requiredPaths: z.array(z.string()),
220
+ allowedPaths: z.array(z.string()),
221
+ observedPaths: z.array(z.string()),
222
+ missingPaths: z.array(z.string()),
223
+ unexpectedPaths: z.array(z.string()),
224
+ enumMismatches: z.array(z.string()),
225
+ }).strict(),
226
+ executionSignature: z.string().min(1).optional(),
181
227
  findings: z.array(z.string()),
182
228
  }).strict();
183
229
  export const backendTestMarkdownPytestCorrespondenceFactsSchema = z.object({
184
- schemaId: z.literal("backend-test-markdown-pytest-correspondence-facts-v2"),
185
- schemaVersion: z.literal(2),
230
+ schemaId: z.literal("backend-test-markdown-pytest-correspondence-facts-v3"),
231
+ schemaVersion: z.literal(3),
186
232
  taskId: z.string().min(1),
187
233
  status: z.enum(["PASS", "FAIL", "UNAVAILABLE"]),
188
234
  inputFiles: z.array(inputFileSchema),
@@ -203,6 +249,10 @@ export const backendTestMarkdownPytestCorrespondenceFactsSchema = z.object({
203
249
  crossCuttingTestPointCount: z.number().int().min(0),
204
250
  unclassifiedTestPointCount: z.number().int().min(0),
205
251
  duplicateBindingTestPointCount: z.number().int().min(0),
252
+ payloadSafeCount: z.number().int().min(0),
253
+ payloadUnsafeCount: z.number().int().min(0),
254
+ payloadUnavailableCount: z.number().int().min(0),
255
+ duplicateExecutionSignatureCount: z.number().int().min(0),
206
256
  }).strict(),
207
257
  entries: z.array(correspondenceEntrySchema),
208
258
  findings: z.array(z.string()),
@@ -236,6 +286,64 @@ function openApiTestPoints(field, dimension, values = []) {
236
286
  return [`${prefix}-VALID-CLASS`, `${prefix}-VALID-MIXED`, `${prefix}-UPPERCASE`, `${prefix}-WHITESPACE`, `${prefix}-PUNCTUATION`, `${prefix}-SLASH`, `${prefix}-EMOJI`, `${prefix}-CONTROL`];
237
287
  return [`${prefix}-PRESENT`, `${prefix}-MISSING`, `${prefix}-NULL`, `${prefix}-WRONG-TYPE`];
238
288
  }
289
+ async function collectSourceCoverageFacts(input) {
290
+ const paths = orderedUnique([
291
+ input.sourceBinding.requirementPath,
292
+ ...input.sourceBinding.referencePaths,
293
+ ]);
294
+ const facts = [];
295
+ const seen = new Set();
296
+ for (const sourcePath of paths) {
297
+ const physicalPath = resolveBoundSourcePath(input.workspaceRoot, input.sourceBinding.taskId, sourcePath);
298
+ let text;
299
+ try {
300
+ text = await readFile(physicalPath, "utf8");
301
+ }
302
+ catch {
303
+ continue;
304
+ }
305
+ const sourceSha256 = createHash("sha256").update(await readFile(physicalPath)).digest("hex");
306
+ const lines = text.split(/\r?\n/);
307
+ const add = (kind, key, lineIndex, quote) => {
308
+ const normalizedQuote = quote.trim().slice(0, 1000);
309
+ if (!normalizedQuote)
310
+ return;
311
+ const identity = `${kind}:${key}:${sourcePath}:${lineIndex + 1}`;
312
+ if (seen.has(identity))
313
+ return;
314
+ seen.add(identity);
315
+ const escaped = key.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
316
+ const covered = kind === "business-rule"
317
+ ? input.generatedMarkdown.includes(normalizedQuote)
318
+ : new RegExp(`\\b${escaped}\\b`).test(input.generatedMarkdown);
319
+ facts.push({
320
+ kind,
321
+ key,
322
+ sourcePath,
323
+ sourceSha256,
324
+ lineStart: lineIndex + 1,
325
+ lineEnd: lineIndex + 1,
326
+ quote: normalizedQuote,
327
+ covered,
328
+ });
329
+ };
330
+ lines.forEach((line, lineIndex) => {
331
+ for (const key of line.match(SOURCE_RULE_ID_IN_TEXT) ?? [])
332
+ add("requirement-id", key, lineIndex, line);
333
+ for (const key of line.match(SOURCE_ERROR_CODE_IN_TEXT) ?? [])
334
+ add("error-code", key, lineIndex, line);
335
+ for (const match of line.matchAll(JSON_FIELD_IN_TEXT))
336
+ add("dto-field", match[1], lineIndex, line);
337
+ for (const match of line.matchAll(DTO_FIELD_IN_TEXT))
338
+ add("dto-field", match[1], lineIndex, line);
339
+ const trimmed = line.trim();
340
+ if (trimmed.length >= 8 && trimmed.length <= 1000 && !/^\s*(?:#|\||[-:]{3,})/.test(trimmed) && NATURAL_BUSINESS_RULE_KEYWORDS.test(trimmed) && NATURAL_BUSINESS_RULE_NORMATIVE.test(trimmed)) {
341
+ add("business-rule", `BR-SOURCE-${sourceSha256.slice(0, 8).toUpperCase()}-L${lineIndex + 1}`, lineIndex, trimmed);
342
+ }
343
+ });
344
+ }
345
+ return facts.sort((left, right) => left.kind.localeCompare(right.kind) || left.key.localeCompare(right.key) || left.sourcePath.localeCompare(right.sourcePath) || left.lineStart - right.lineStart);
346
+ }
239
347
  function resolveBoundSourcePath(workspaceRoot, taskId, sourcePath) {
240
348
  if (path.isAbsolute(sourcePath))
241
349
  return sourcePath;
@@ -385,8 +493,14 @@ function sectionBody(body, aliases) {
385
493
  }
386
494
  function listTokens(body, aliases, pattern) {
387
495
  const section = sectionBody(body, aliases).replaceAll("`", "");
388
- const source = pattern.source.replace(/^\^/, "").replace(/\$$/, "");
389
- return orderedUnique(section.match(new RegExp(source, "g")) ?? []);
496
+ // Prefer dedicated token patterns; never strip ^/$ from alternation-based anchors
497
+ // (that would turn `^A$|^B$` into `A$|^B` and break scanning).
498
+ const tokenSource = pattern === RULE_KEY
499
+ ? RULE_KEY_TOKEN.source
500
+ : pattern === TEST_POINT
501
+ ? TEST_POINT.source.replace(/^\^/, "").replace(/\$$/, "")
502
+ : pattern.source.replace(/^\^/, "").replace(/\$$/, "");
503
+ return orderedUnique(section.match(new RegExp(tokenSource, "g")) ?? []);
390
504
  }
391
505
  function scenarioTypes(body) {
392
506
  return orderedUnique(sectionBody(body, CASE_SECTION_ALIASES.scenarioTypes)
@@ -423,8 +537,40 @@ function testPointBindingFacts(body, testPoints) {
423
537
  }
424
538
  function declaredPrimarySymbols(body) {
425
539
  const value = automationLabelBody(body, ["primary symbol", "Primary Symbol", "主测试符号"]);
540
+ if (/^(?:none|无|n\/a|not-applicable)$/i.test(value.trim()))
541
+ return [];
426
542
  return orderedUnique(value.match(/\btest_[A-Za-z0-9_]+\b/g) ?? []);
427
543
  }
544
+ function isEvidenceOnlyMarkdownCase(testCase) {
545
+ const hasVariant = testCase.testPointBindings.some((item) => item.mode === "variant");
546
+ const hasNonVariant = testCase.testPointBindings.some((item) => item.mode === "assertion" || item.mode === "cross-cutting");
547
+ return (testCase.declaredScripts.length === 0 &&
548
+ testCase.declaredPrimarySymbols.length === 0 &&
549
+ !hasVariant &&
550
+ hasNonVariant);
551
+ }
552
+ function payloadContract(body) {
553
+ const contractValue = automationLabelBody(body, ["Payload Contract", "载荷合同"]);
554
+ const requiredValue = automationLabelBody(body, ["Payload Required Paths", "载荷必需路径"]);
555
+ const allowedValue = automationLabelBody(body, ["Payload Allowed Paths", "载荷允许路径"]);
556
+ const enumValue = automationLabelBody(body, ["Payload Enum", "载荷枚举"]);
557
+ const tokens = (value) => orderedUnique(value.split(/[,;,;]/).map((item) => item.trim()).filter((item) => /^[A-Za-z_][A-Za-z0-9_.]*$/.test(item)));
558
+ const enumValues = {};
559
+ for (const assignment of enumValue.split(/[,;,;]/).map((item) => item.trim()).filter(Boolean)) {
560
+ const match = /^([A-Za-z_][A-Za-z0-9_.]*)\s*=\s*(.+)$/.exec(assignment);
561
+ if (!match)
562
+ continue;
563
+ enumValues[match[1]] = orderedUnique(match[2].split(/\||\//).map((item) => item.trim().replace(/^['"]|['"]$/g, "")).filter(Boolean));
564
+ }
565
+ const noPayload = /^(?:none|无|not-applicable|n\/a)$/i.test(contractValue);
566
+ return {
567
+ declared: noPayload || requiredValue.length > 0 || allowedValue.length > 0 || enumValue.length > 0,
568
+ noPayload,
569
+ requiredPaths: tokens(requiredValue),
570
+ allowedPaths: tokens(allowedValue),
571
+ enumValues,
572
+ };
573
+ }
428
574
  function cleanTitle(heading, id) {
429
575
  return heading.replace(/^##\s+/, "").replace(/^BE-[A-Z0-9_-]+-\d{2,3}\s*(?:[||—–-]\s*)?/i, "").trim() || id;
430
576
  }
@@ -466,6 +612,7 @@ async function parseMarkdownCases(workspaceRoot) {
466
612
  scenarioTypes: scenarioTypes(body),
467
613
  declaredScripts: declaredScripts(body),
468
614
  declaredPrimarySymbols: declaredPrimarySymbols(body),
615
+ payloadContract: payloadContract(body),
469
616
  body,
470
617
  });
471
618
  }
@@ -483,8 +630,17 @@ function parseTableRow(line) {
483
630
  index += 1;
484
631
  }
485
632
  else if (char === "|") {
486
- cells.push(current.trim());
487
- current = "";
633
+ const prev = trimmed[index - 1] ?? "";
634
+ const next = trimmed[index + 1] ?? "";
635
+ // In-cell enum/value pipes look like ACTIVE|ARCHIVED (alnum on both sides).
636
+ // Column separators use " | " or touch markdown separator dashes (---|---).
637
+ if (/[A-Za-z0-9_]/.test(prev) && /[A-Za-z0-9_]/.test(next)) {
638
+ current += "|";
639
+ }
640
+ else {
641
+ cells.push(current.trim());
642
+ current = "";
643
+ }
488
644
  }
489
645
  else {
490
646
  current += char;
@@ -687,7 +843,10 @@ export async function analyzeBackendTestCaseCoverage(input) {
687
843
  if (!rowCaseIds.includes(testCase.caseId))
688
844
  findings.push(`${testCase.caseId} references ${ruleKey} but is absent from its Coverage Matrix Case IDs`);
689
845
  }
690
- let status = declaredStatus;
846
+ const normalizedDeclaredStatus = /^(?:COVERED|PARTIAL|GAP|CONFLICT)$/.test(declaredStatus)
847
+ ? declaredStatus
848
+ : "PARTIAL";
849
+ let status = normalizedDeclaredStatus;
691
850
  if (status === "COVERED" && missingTestPoints.length > 0)
692
851
  status = "PARTIAL";
693
852
  if (missingTestPoints.length > 0)
@@ -707,7 +866,7 @@ export async function analyzeBackendTestCaseCoverage(input) {
707
866
  rule: row.Rule || "unavailable",
708
867
  requiredTestPoints,
709
868
  caseIds: rowCaseIds,
710
- declaredStatus: /^(?:COVERED|PARTIAL|GAP|CONFLICT)$/.test(declaredStatus) ? declaredStatus : "PARTIAL",
869
+ declaredStatus: normalizedDeclaredStatus,
711
870
  coveredTestPoints,
712
871
  missingTestPoints,
713
872
  status,
@@ -749,8 +908,12 @@ export async function analyzeBackendTestCaseCoverage(input) {
749
908
  findings.push(`${testCase.caseId} missing section or entries: 测试点`);
750
909
  if (testCase.scenarioTypes.length === 0)
751
910
  findings.push(`${testCase.caseId} missing section or entries: 场景类型`);
752
- if (testCase.declaredPrimarySymbols.length !== 1)
911
+ if (isEvidenceOnlyMarkdownCase(testCase)) {
912
+ // Evidence-only meta cases intentionally declare 脚本/primary symbol as 无 and only keep assertion/cross-cutting TPs.
913
+ }
914
+ else if (testCase.declaredPrimarySymbols.length !== 1) {
753
915
  findings.push(`${testCase.caseId} must declare exactly one primary symbol`);
916
+ }
754
917
  if (testCase.unclassifiedTestPoints.length > 0)
755
918
  findings.push(`${testCase.caseId} has unclassified Test Points: ${testCase.unclassifiedTestPoints.join(", ")}`);
756
919
  if (testCase.duplicateBindingTestPoints.length > 0)
@@ -772,8 +935,28 @@ export async function analyzeBackendTestCaseCoverage(input) {
772
935
  if (uniqueOwners.length > 1)
773
936
  findings.push(`${testPoint} is bound as a non-cross-cutting Test Point by multiple Cases: ${uniqueOwners.join(", ")}`);
774
937
  }
775
- const explicitAc = input.sourceBinding.requirementIds.filter((id) => id.startsWith("AC-"));
938
+ const generatedMarkdown = [readme, ...cases.map((item) => item.body)].join("\n\n");
939
+ const sourceCoverageFacts = await collectSourceCoverageFacts({
940
+ workspaceRoot: input.workspaceRoot,
941
+ sourceBinding: input.sourceBinding,
942
+ generatedMarkdown,
943
+ });
944
+ const residualSourceFacts = sourceCoverageFacts.filter((fact) => !fact.covered);
945
+ for (const fact of residualSourceFacts) {
946
+ findings.push(`source coverage is missing ${fact.kind} ${fact.key} from ${fact.sourcePath}`);
947
+ }
948
+ const sourceRequirementIds = sourceCoverageFacts
949
+ .filter((fact) => fact.kind === "requirement-id")
950
+ .map((fact) => fact.key);
951
+ const explicitAc = orderedUnique([
952
+ ...input.sourceBinding.requirementIds.filter((id) => id.startsWith("AC-")),
953
+ ...sourceRequirementIds.filter((id) => id.startsWith("AC-")),
954
+ ]);
776
955
  const coveredAc = new Set(cases.flatMap((item) => item.acIds));
956
+ for (const acId of explicitAc) {
957
+ if (!coveredAc.has(acId))
958
+ findings.push(`explicit source AC is missing from Markdown Cases: ${acId}`);
959
+ }
777
960
  const [enumValueCount, coveredEnumValueCount] = dimensionCounts(rules, /enum/);
778
961
  const [boundaryPointCount, coveredBoundaryPointCount] = dimensionCounts(rules, /boundary|length|numeric/);
779
962
  const [formatClassCount, coveredFormatClassCount] = dimensionCounts(rules, /format|pattern|charset/);
@@ -782,8 +965,8 @@ export async function analyzeBackendTestCaseCoverage(input) {
782
965
  const invalidEquivalenceClassCount = rules.reduce((sum, rule) => sum + rule.requiredTestPoints.filter((point) => invalidPointPattern.test(point)).length, 0);
783
966
  const coveredInvalidEquivalenceClassCount = rules.reduce((sum, rule) => sum + rule.coveredTestPoints.filter((point) => invalidPointPattern.test(point)).length, 0);
784
967
  const facts = backendTestCaseCoverageFactsSchema.parse({
785
- schemaId: "backend-test-case-coverage-facts-v3",
786
- schemaVersion: 3,
968
+ schemaId: "backend-test-case-coverage-facts-v4",
969
+ schemaVersion: 4,
787
970
  taskId: input.sourceBinding.taskId,
788
971
  status: findings.length === 0 ? "PASS" : "FAIL",
789
972
  sourceBinding: input.sourceBinding,
@@ -796,6 +979,13 @@ export async function analyzeBackendTestCaseCoverage(input) {
796
979
  }))).filter(Boolean),
797
980
  ]),
798
981
  coverageScope: coverageScope.scope,
982
+ sourceCoverage: {
983
+ sourceFactCount: sourceCoverageFacts.length,
984
+ coveredSourceFactCount: sourceCoverageFacts.length - residualSourceFacts.length,
985
+ residualSourceFactCount: residualSourceFacts.length,
986
+ facts: sourceCoverageFacts,
987
+ residual: residualSourceFacts,
988
+ },
799
989
  summary: {
800
990
  explicitAcCount: explicitAc.length,
801
991
  coveredAcCount: explicitAc.filter((id) => coveredAc.has(id)).length,
@@ -846,6 +1036,22 @@ export async function analyzeBackendTestCaseCoverage(input) {
846
1036
  `- Scope Evidence: ${facts.coverageScope.scopeEvidence.join(", ") || "Unavailable"}`,
847
1037
  ...(facts.coverageScope.completenessClaim === "affected-scope" ? ["- Coverage is scoped to affected operations; it is not whole-API completeness."] : ["- Coverage fully evaluates the declared affected operations; it is not whole-API completeness unless every operation is explicitly listed."]),
848
1038
  "",
1039
+ "## Source Coverage Inventory",
1040
+ "",
1041
+ `- Source facts: ${facts.sourceCoverage.sourceFactCount}`,
1042
+ `- Covered source facts: ${facts.sourceCoverage.coveredSourceFactCount}`,
1043
+ `- Residual source facts: ${facts.sourceCoverage.residualSourceFactCount}`,
1044
+ "",
1045
+ "| Kind | Key | Source Anchor | Source SHA-256 | Exact Quote | Status |",
1046
+ "|---|---|---|---|---|---|",
1047
+ ...facts.sourceCoverage.facts.map((fact) => `| ${fact.kind} | ${fact.key} | ${fact.sourcePath}:L${fact.lineStart} | ${fact.sourceSha256} | ${fact.quote.replaceAll("|", "\\|")} | ${fact.covered ? "COVERED" : "MISSING_AFTER_SYNC"} |`),
1048
+ "",
1049
+ "## Residual Missing Source Scenarios",
1050
+ "",
1051
+ ...(facts.sourceCoverage.residual.length > 0
1052
+ ? facts.sourceCoverage.residual.map((fact) => `- ${fact.kind} \`${fact.key}\` from \`${fact.sourcePath}:L${fact.lineStart}\` (sha256 \`${fact.sourceSha256}\`), quote: “${fact.quote}”: missing after the bounded Markdown synchronization round; no full-coverage claim is allowed.`)
1053
+ : ["- None"]),
1054
+ "",
849
1055
  "## Coverage Summary",
850
1056
  "",
851
1057
  "| Dimension | Total | Covered |",
@@ -909,6 +1115,103 @@ function symbolCaseId(symbol) {
909
1115
  const match = symbol.match(/^test_(BE(?:_[A-Z0-9]+)+?_\d{2,3})(?:_|$)/i);
910
1116
  return match ? canonicalCaseId(match[1].replaceAll("_", "-")) : undefined;
911
1117
  }
1118
+ function pythonPayloadShapes(source) {
1119
+ const program = String.raw `
1120
+ import ast,json,sys
1121
+ source=sys.stdin.buffer.read().decode('utf-8')
1122
+ tree=ast.parse(source)
1123
+ GLOBAL_PAYLOADS={}
1124
+ for top in tree.body:
1125
+ if isinstance(top,ast.Assign):
1126
+ names=[target.id for target in top.targets if isinstance(target,ast.Name)]
1127
+ candidates=[]
1128
+ for node in ast.walk(top.value):
1129
+ if isinstance(node,ast.Call) and isinstance(node.func,ast.Attribute) and node.func.attr == 'param' and node.args and isinstance(node.args[0],ast.Dict): candidates.append(node.args[0])
1130
+ if isinstance(top.value,ast.Dict) and any(name.lower() in {'payload','body','request_body','request_json','dto'} for name in names): candidates.append(top.value)
1131
+ for name in names: GLOBAL_PAYLOADS[name]=candidates
1132
+ FUNCTION_PAYLOADS={}
1133
+ for top in tree.body:
1134
+ if isinstance(top,(ast.FunctionDef,ast.AsyncFunctionDef)) and not top.name.startswith('test_'):
1135
+ candidates=[]
1136
+ assigned={}
1137
+ for node in ast.walk(top):
1138
+ if isinstance(node,ast.Assign) and isinstance(node.value,ast.Dict):
1139
+ for target in node.targets:
1140
+ if isinstance(target,ast.Name): assigned[target.id]=node.value
1141
+ if isinstance(node,ast.Return):
1142
+ if isinstance(node.value,ast.Dict): candidates.append(node.value)
1143
+ elif isinstance(node.value,ast.Name) and node.value.id in assigned: candidates.append(assigned[node.value.id])
1144
+ if candidates: FUNCTION_PAYLOADS[top.name]=candidates
1145
+
1146
+ def literal(node):
1147
+ if isinstance(node, ast.Constant) and isinstance(node.value, (str,int,float,bool,type(None))):
1148
+ return repr(node.value)
1149
+ return None
1150
+
1151
+ def collect_dict(node, prefix, paths, values):
1152
+ if not isinstance(node, ast.Dict): return
1153
+ for key,value in zip(node.keys,node.values):
1154
+ if not isinstance(key, ast.Constant) or not isinstance(key.value,str): continue
1155
+ current = key.value if not prefix else prefix + '.' + key.value
1156
+ paths.add(current)
1157
+ lv=literal(value)
1158
+ if lv is not None: values.setdefault(current,set()).add(lv)
1159
+ if isinstance(value, ast.Dict): collect_dict(value,current,paths,values)
1160
+ elif isinstance(value,(ast.List,ast.Tuple)):
1161
+ for item in value.elts:
1162
+ if isinstance(item,ast.Dict): collect_dict(item,current,paths,values)
1163
+
1164
+ def payload_nodes(fn):
1165
+ result=[]
1166
+ for dec in fn.decorator_list:
1167
+ for node in ast.walk(dec):
1168
+ if isinstance(node,ast.Call) and isinstance(node.func,ast.Attribute) and node.func.attr == 'param' and node.args:
1169
+ if isinstance(node.args[0],ast.Dict): result.append(node.args[0])
1170
+ elif isinstance(node,ast.Name) and node.id in GLOBAL_PAYLOADS:
1171
+ result.extend(GLOBAL_PAYLOADS[node.id])
1172
+ for node in ast.walk(fn):
1173
+ if isinstance(node,ast.Assign):
1174
+ names=[target.id for target in node.targets if isinstance(target,ast.Name)]
1175
+ if any(name.lower() in {'payload','body','request_body','request_json','dto'} for name in names) and isinstance(node.value,ast.Dict): result.append(node.value)
1176
+ if isinstance(node,ast.Call):
1177
+ if isinstance(node.func,ast.Name) and node.func.id in FUNCTION_PAYLOADS:
1178
+ result.extend(FUNCTION_PAYLOADS[node.func.id])
1179
+ override_keys=[]; override_values=[]
1180
+ for keyword in node.keywords:
1181
+ if keyword.arg:
1182
+ override_keys.append(ast.Constant(keyword.arg)); override_values.append(keyword.value)
1183
+ if override_keys: result.append(ast.Dict(keys=override_keys,values=override_values))
1184
+ for keyword in node.keywords:
1185
+ if keyword.arg in {'json','data','payload','body','request_body'} and isinstance(keyword.value,ast.Dict): result.append(keyword.value)
1186
+ return result
1187
+
1188
+ out={}
1189
+ for fn in ast.walk(tree):
1190
+ if isinstance(fn,(ast.FunctionDef,ast.AsyncFunctionDef)) and fn.name.startswith('test_'):
1191
+ paths=set(); values={}
1192
+ for node in payload_nodes(fn): collect_dict(node,'',paths,values)
1193
+ out[fn.name]={'paths':sorted(paths),'values':{key:sorted(items) for key,items in sorted(values.items())}}
1194
+ print(json.dumps(out,ensure_ascii=True))
1195
+ `;
1196
+ // Pass bytes and decode stdin explicitly as UTF-8 in Python. Clean-env
1197
+ // Windows Python may expose stdin as GBK; decoding generated Chinese/emoji
1198
+ // source through that locale can create surrogateescape code points and make
1199
+ // ast.parse fail with "surrogates not allowed".
1200
+ const result = spawnSync("python", ["-c", program], { input: Buffer.from(source, "utf8"), encoding: "utf8", maxBuffer: 4 * 1024 * 1024, windowsHide: true });
1201
+ if (result.status !== 0 || !result.stdout.trim())
1202
+ return new Map();
1203
+ try {
1204
+ const parsed = JSON.parse(result.stdout);
1205
+ return new Map(Object.entries(parsed));
1206
+ }
1207
+ catch {
1208
+ return new Map();
1209
+ }
1210
+ }
1211
+ function metadataCaseIds(region) {
1212
+ return orderedUnique([...region.matchAll(/^\s*Case-ID\s*:\s*(BE-[A-Z0-9]+(?:-[A-Z0-9]+)*)\s*$/gmi)]
1213
+ .map((match) => canonicalCaseId(match[1])));
1214
+ }
912
1215
  function metadataTestPoints(region, label) {
913
1216
  const escaped = label.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
914
1217
  const match = new RegExp(`^\\s*${escaped}\\s*:\\s*(.*)$`, "mi").exec(region);
@@ -928,6 +1231,27 @@ function pytestFunctionRegionStart(source, functionIndex) {
928
1231
  }
929
1232
  return regionStart;
930
1233
  }
1234
+ function pytestFunctionBodyEnd(source, signatureEnd, functionIndent) {
1235
+ let cursor = source.indexOf("\n", signatureEnd);
1236
+ if (cursor < 0)
1237
+ return source.length;
1238
+ cursor += 1;
1239
+ const baseIndent = functionIndent.replaceAll("\t", " ").length;
1240
+ while (cursor < source.length) {
1241
+ const nextLine = source.indexOf("\n", cursor);
1242
+ const lineEnd = nextLine < 0 ? source.length : nextLine;
1243
+ const line = source.slice(cursor, lineEnd).replace(/\r$/, "");
1244
+ if (line.trim()) {
1245
+ const indentation = line.match(/^[ \t]*/)?.[0]?.replaceAll("\t", " ").length ?? 0;
1246
+ if (indentation <= baseIndent)
1247
+ return cursor;
1248
+ }
1249
+ if (nextLine < 0)
1250
+ return source.length;
1251
+ cursor = nextLine + 1;
1252
+ }
1253
+ return source.length;
1254
+ }
931
1255
  function pytestParameterCollections(source) {
932
1256
  const collections = new Map();
933
1257
  for (const match of source.matchAll(/^([A-Z][A-Z0-9_]*)\s*=\s*\[([\s\S]*?)^\]/gm)) {
@@ -939,12 +1263,13 @@ function pytestSymbols(script, source) {
939
1263
  const matches = [...source.matchAll(/^([ \t]*)(?:async\s+)?def\s+(test_[A-Za-z0-9_]+)\s*\([^)]*\)\s*(?:->\s*[^:\r\n]+)?\s*:/gm)];
940
1264
  const regionStarts = matches.map((match) => pytestFunctionRegionStart(source, match.index));
941
1265
  const parameterCollections = pytestParameterCollections(source);
1266
+ const payloadShapes = pythonPayloadShapes(source);
942
1267
  return matches.map((match, index) => {
943
1268
  const regionStart = regionStarts[index];
944
- const regionEnd = regionStarts[index + 1] ?? source.length;
1269
+ const regionEnd = pytestFunctionBodyEnd(source, match.index + match[0].length, match[1]);
945
1270
  const region = source.slice(match.index, regionEnd);
946
1271
  const decorators = source.slice(regionStart, match.index);
947
- const ids = caseIds(region);
1272
+ const ids = metadataCaseIds(region);
948
1273
  const fromSymbol = symbolCaseId(match[2]);
949
1274
  if (fromSymbol)
950
1275
  ids.unshift(fromSymbol);
@@ -954,6 +1279,7 @@ function pytestSymbols(script, source) {
954
1279
  ]);
955
1280
  const assertionTestPoints = metadataTestPoints(region, "Assertion-Test-Points");
956
1281
  const crossCuttingTestPoints = metadataTestPoints(region, "Cross-Cutting-Test-Points");
1282
+ const payloadShape = payloadShapes.get(match[2]) ?? { paths: [], values: {} };
957
1283
  return {
958
1284
  script,
959
1285
  symbol: match[2],
@@ -962,9 +1288,51 @@ function pytestSymbols(script, source) {
962
1288
  parameterIds,
963
1289
  assertionTestPoints,
964
1290
  crossCuttingTestPoints,
1291
+ payloadPaths: payloadShape.paths,
1292
+ payloadValues: payloadShape.values,
965
1293
  };
966
1294
  });
967
1295
  }
1296
+ function executionSignature(testCase) {
1297
+ const operation = (testCase.body.match(/\b(?:GET|POST|PUT|PATCH|DELETE)\s+\/[^\s`|)]+/i)?.[0] ?? "operation-unavailable").toUpperCase();
1298
+ const inputPartition = [sectionBody(testCase.body, ["测试数据", "Test Data"]), testCase.scenarioTypes.join(" "), testCase.testPoints.join(" ")].join(" ");
1299
+ const precondition = sectionBody(testCase.body, ["前置条件", "Preconditions", "Precondition"]);
1300
+ const expectedOutcome = sectionBody(testCase.body, ["预期结果", "Expected Results", "Expected Result"]);
1301
+ const normalized = [operation, inputPartition, precondition, expectedOutcome]
1302
+ .map((value) => value.replaceAll(/\s+/g, " ").trim().toLowerCase())
1303
+ .join("\n");
1304
+ return createHash("sha256").update(normalized).digest("hex");
1305
+ }
1306
+ function assessPayloadContract(testCase, refs) {
1307
+ const observedPaths = orderedUnique(refs.flatMap((item) => item.payloadPaths));
1308
+ const requiredPaths = testCase.payloadContract.requiredPaths;
1309
+ const allowedPaths = testCase.payloadContract.allowedPaths;
1310
+ if (!testCase.payloadContract.declared) {
1311
+ return { status: "UNAVAILABLE", requiredPaths, allowedPaths, observedPaths, missingPaths: [], unexpectedPaths: [], enumMismatches: [] };
1312
+ }
1313
+ const missingPaths = requiredPaths.filter((item) => !observedPaths.includes(item));
1314
+ const unexpectedPaths = testCase.payloadContract.noPayload
1315
+ ? observedPaths
1316
+ : allowedPaths.length > 0
1317
+ ? observedPaths.filter((item) => !allowedPaths.includes(item) && !allowedPaths.some((allowed) => item.startsWith(`${allowed}.`)))
1318
+ : [];
1319
+ const enumMismatches = [];
1320
+ for (const [field, allowed] of Object.entries(testCase.payloadContract.enumValues)) {
1321
+ const observed = orderedUnique(refs.flatMap((item) => item.payloadValues[field] ?? []).map((value) => value.replace(/^['"]|['"]$/g, "")));
1322
+ for (const value of observed)
1323
+ if (!allowed.includes(value))
1324
+ enumMismatches.push(`${field}=${value}; allowed=${allowed.join("|")}`);
1325
+ }
1326
+ return {
1327
+ status: missingPaths.length > 0 || unexpectedPaths.length > 0 || enumMismatches.length > 0 ? "UNSAFE" : "SAFE",
1328
+ requiredPaths,
1329
+ allowedPaths,
1330
+ observedPaths,
1331
+ missingPaths,
1332
+ unexpectedPaths,
1333
+ enumMismatches,
1334
+ };
1335
+ }
968
1336
  export async function analyzeBackendTestMarkdownPytestCorrespondence(input) {
969
1337
  const cases = await parseMarkdownCases(input.workspaceRoot);
970
1338
  const modulePaths = orderedUnique(cases.map((item) => item.markdownPath));
@@ -998,21 +1366,44 @@ export async function analyzeBackendTestMarkdownPytestCorrespondence(input) {
998
1366
  ...crossCuttingTestPoints.filter((point) => crossCuttingBindings.includes(point)),
999
1367
  ]);
1000
1368
  const declaredPrimarySymbol = testCase.declaredPrimarySymbols[0];
1369
+ const evidenceOnly = isEvidenceOnlyMarkdownCase(testCase);
1001
1370
  const cardinality = refs.length === 0 ? "1:0" : refs.length === 1 ? "1:1" : "1:N";
1002
- let status = refs.length === 0 ? "MISSING_PYTEST" : refs.length > 1 ? "MULTIPLE_PYTEST" : "EXACT_1_TO_1";
1371
+ let status = evidenceOnly && refs.length === 0
1372
+ ? "EXACT_1_TO_1"
1373
+ : refs.length === 0
1374
+ ? "MISSING_PYTEST"
1375
+ : refs.length > 1
1376
+ ? "MULTIPLE_PYTEST"
1377
+ : "EXACT_1_TO_1";
1003
1378
  const entryFindings = [];
1004
- if (refs.length === 0)
1005
- entryFindings.push(`${testCase.caseId} has no pytest symbol`);
1006
- if (refs.length > 1)
1007
- entryFindings.push(`${testCase.caseId} maps to multiple pytest symbols: ${symbols.join(", ")}`);
1008
- if (!declared || declared !== expected || actualScripts.some((script) => script !== expected) || !declaredPrimarySymbol || symbols.some((symbol) => symbol !== declaredPrimarySymbol)) {
1009
- if (refs.length > 0 && status === "EXACT_1_TO_1")
1379
+ const payloadAssessment = assessPayloadContract(testCase, refs);
1380
+ const signature = executionSignature(testCase);
1381
+ if (evidenceOnly) {
1382
+ if (refs.length > 0) {
1010
1383
  status = "SCRIPT_MISMATCH";
1011
- entryFindings.push(`${testCase.caseId} script/primary-symbol mapping mismatch: declaredScript=${declared ?? "none"}, expectedScript=${expected}, actual=${actualScripts.join(", ") || "none"}, declaredPrimarySymbol=${declaredPrimarySymbol ?? "none"}, actualSymbols=${symbols.join(", ") || "none"}`);
1384
+ entryFindings.push(`${testCase.caseId} evidence-only Markdown Case must not materialize a business pytest symbol; actualSymbols=${symbols.join(", ") || "none"}`);
1385
+ }
1012
1386
  }
1013
- const missingVariant = variantTestPoints.filter((point) => !parameterIds.includes(point));
1014
- const missingAssertion = assertionTestPoints.filter((point) => !assertionBindings.includes(point));
1015
- const missingCrossCutting = crossCuttingTestPoints.filter((point) => !crossCuttingBindings.includes(point));
1387
+ else {
1388
+ if (refs.length === 0)
1389
+ entryFindings.push(`${testCase.caseId} has no pytest symbol`);
1390
+ if (refs.length > 1)
1391
+ entryFindings.push(`${testCase.caseId} maps to multiple pytest symbols: ${symbols.join(", ")}`);
1392
+ if (!declared || declared !== expected || actualScripts.some((script) => script !== expected) || !declaredPrimarySymbol || symbols.some((symbol) => symbol !== declaredPrimarySymbol)) {
1393
+ if (refs.length > 0 && status === "EXACT_1_TO_1")
1394
+ status = "SCRIPT_MISMATCH";
1395
+ entryFindings.push(`${testCase.caseId} script/primary-symbol mapping mismatch: declaredScript=${declared ?? "none"}, expectedScript=${expected}, actual=${actualScripts.join(", ") || "none"}, declaredPrimarySymbol=${declaredPrimarySymbol ?? "none"}, actualSymbols=${symbols.join(", ") || "none"}`);
1396
+ }
1397
+ }
1398
+ const missingVariant = evidenceOnly
1399
+ ? []
1400
+ : variantTestPoints.filter((point) => !parameterIds.includes(point));
1401
+ const missingAssertion = evidenceOnly
1402
+ ? []
1403
+ : assertionTestPoints.filter((point) => !assertionBindings.includes(point));
1404
+ const missingCrossCutting = evidenceOnly
1405
+ ? []
1406
+ : crossCuttingTestPoints.filter((point) => !crossCuttingBindings.includes(point));
1016
1407
  const extraPoints = orderedUnique([...parameterIds, ...assertionBindings, ...crossCuttingBindings]).filter((point) => !testCase.testPoints.includes(point));
1017
1408
  if (testCase.unclassifiedTestPoints.length > 0 || testCase.duplicateBindingTestPoints.length > 0) {
1018
1409
  if (refs.length > 0 && status === "EXACT_1_TO_1")
@@ -1039,6 +1430,16 @@ export async function analyzeBackendTestMarkdownPytestCorrespondence(input) {
1039
1430
  status = "TEST_POINT_BINDING_EXTRA";
1040
1431
  entryFindings.push(`${testCase.caseId} has extra pytest Test Point bindings: ${extraPoints.join(", ")}`);
1041
1432
  }
1433
+ if (payloadAssessment.status === "UNAVAILABLE") {
1434
+ if (refs.length > 0 && status === "EXACT_1_TO_1")
1435
+ status = "PAYLOAD_CONTRACT_UNAVAILABLE";
1436
+ entryFindings.push(`${testCase.caseId} has no deterministic Payload Contract declaration`);
1437
+ }
1438
+ else if (payloadAssessment.status === "UNSAFE") {
1439
+ if (refs.length > 0 && status === "EXACT_1_TO_1")
1440
+ status = "PAYLOAD_SHAPE_MISMATCH";
1441
+ entryFindings.push(`${testCase.caseId} payload contract mismatch: missing=${payloadAssessment.missingPaths.join(", ") || "none"}; unexpected=${payloadAssessment.unexpectedPaths.join(", ") || "none"}; enum=${payloadAssessment.enumMismatches.join(", ") || "none"}`);
1442
+ }
1042
1443
  findings.push(...entryFindings);
1043
1444
  entries.push({
1044
1445
  markdownModule: testCase.markdownPath,
@@ -1056,6 +1457,8 @@ export async function analyzeBackendTestMarkdownPytestCorrespondence(input) {
1056
1457
  parameterIds,
1057
1458
  cardinality,
1058
1459
  status,
1460
+ payloadAssessment,
1461
+ executionSignature: signature,
1059
1462
  findings: entryFindings,
1060
1463
  });
1061
1464
  }
@@ -1082,9 +1485,27 @@ export async function analyzeBackendTestMarkdownPytestCorrespondence(input) {
1082
1485
  parameterIds: symbol.parameterIds,
1083
1486
  cardinality: "0:1",
1084
1487
  status: unknownId ? "EXTRA_PYTEST" : "SYMBOL_MISSING_CASE_ID",
1488
+ payloadAssessment: {
1489
+ status: "UNAVAILABLE",
1490
+ requiredPaths: [],
1491
+ allowedPaths: [],
1492
+ observedPaths: symbol.payloadPaths,
1493
+ missingPaths: [],
1494
+ unexpectedPaths: [],
1495
+ enumMismatches: [],
1496
+ },
1085
1497
  findings: [message],
1086
1498
  });
1087
1499
  }
1500
+ const duplicateSignatures = new Map();
1501
+ for (const entry of entries.filter((item) => item.caseId && item.executionSignature)) {
1502
+ const owners = duplicateSignatures.get(entry.executionSignature) ?? [];
1503
+ owners.push(entry.caseId);
1504
+ duplicateSignatures.set(entry.executionSignature, owners);
1505
+ }
1506
+ const duplicateSignatureGroups = [...duplicateSignatures.entries()].filter(([, owners]) => orderedUnique(owners).length > 1);
1507
+ for (const [signature, owners] of duplicateSignatureGroups)
1508
+ findings.push(`duplicate execution signature ${signature}: ${orderedUnique(owners).join(", ")}`);
1088
1509
  const exactModuleCount = modulePaths.filter((modulePath) => {
1089
1510
  const expected = expectedScript(modulePath);
1090
1511
  const moduleCases = cases.filter((item) => item.markdownPath === modulePath);
@@ -1092,8 +1513,8 @@ export async function analyzeBackendTestMarkdownPytestCorrespondence(input) {
1092
1513
  }).length;
1093
1514
  const caseEntries = entries.filter((entry) => entry.cardinality !== "0:1");
1094
1515
  const facts = backendTestMarkdownPytestCorrespondenceFactsSchema.parse({
1095
- schemaId: "backend-test-markdown-pytest-correspondence-facts-v2",
1096
- schemaVersion: 2,
1516
+ schemaId: "backend-test-markdown-pytest-correspondence-facts-v3",
1517
+ schemaVersion: 3,
1097
1518
  taskId: input.taskId,
1098
1519
  status: findings.length === 0 ? "PASS" : "FAIL",
1099
1520
  inputFiles: await inputFileFacts(input.workspaceRoot, [
@@ -1118,6 +1539,10 @@ export async function analyzeBackendTestMarkdownPytestCorrespondence(input) {
1118
1539
  crossCuttingTestPointCount: cases.reduce((sum, item) => sum + item.testPointBindings.filter((binding) => binding.mode === "cross-cutting").length, 0),
1119
1540
  unclassifiedTestPointCount: cases.reduce((sum, item) => sum + item.unclassifiedTestPoints.length, 0),
1120
1541
  duplicateBindingTestPointCount: cases.reduce((sum, item) => sum + item.duplicateBindingTestPoints.length, 0),
1542
+ payloadSafeCount: caseEntries.filter((entry) => entry.payloadAssessment.status === "SAFE").length,
1543
+ payloadUnsafeCount: caseEntries.filter((entry) => entry.payloadAssessment.status === "UNSAFE").length,
1544
+ payloadUnavailableCount: caseEntries.filter((entry) => entry.payloadAssessment.status === "UNAVAILABLE").length,
1545
+ duplicateExecutionSignatureCount: duplicateSignatureGroups.length,
1121
1546
  },
1122
1547
  entries,
1123
1548
  findings: orderedUnique(findings),
@@ -1145,12 +1570,16 @@ export async function analyzeBackendTestMarkdownPytestCorrespondence(input) {
1145
1570
  `- Cross-Cutting Test Points: ${facts.summary.crossCuttingTestPointCount}`,
1146
1571
  `- Unclassified Test Points: ${facts.summary.unclassifiedTestPointCount}`,
1147
1572
  `- Duplicate Test Point Bindings: ${facts.summary.duplicateBindingTestPointCount}`,
1573
+ `- Payload SAFE: ${facts.summary.payloadSafeCount}`,
1574
+ `- Payload UNSAFE: ${facts.summary.payloadUnsafeCount}`,
1575
+ `- Payload UNAVAILABLE: ${facts.summary.payloadUnavailableCount}`,
1576
+ `- Duplicate Execution Signatures: ${facts.summary.duplicateExecutionSignatureCount}`,
1148
1577
  "",
1149
1578
  "## Correspondence Matrix",
1150
1579
  "",
1151
- "| Markdown Module | Case ID | Variant | Assertion | Cross-Cutting | Mapped | Declared Script | Primary Symbol | Actual Script | Pytest Symbol | Parameter IDs | Cardinality | Status |",
1152
- "|---|---|---|---|---|---|---|---|---|---|---|---|---|",
1153
- ...facts.entries.map((entry) => `| ${entry.markdownModule} | ${entry.caseId ?? "—"} | ${entry.variantTestPoints.join(", ") || "—"} | ${entry.assertionTestPoints.join(", ") || "—"} | ${entry.crossCuttingTestPoints.join(", ") || "—"} | ${entry.mappedTestPoints.join(", ") || "—"} | ${entry.declaredScript ?? "—"} | ${entry.declaredPrimarySymbol ?? "—"} | ${entry.actualScripts.join(", ") || "—"} | ${entry.pytestSymbols.join(", ") || "—"} | ${entry.parameterIds.join(", ") || "—"} | ${entry.cardinality} | ${entry.status} |`),
1580
+ "| Markdown Module | Case ID | Variant | Assertion | Cross-Cutting | Mapped | Declared Script | Primary Symbol | Actual Script | Pytest Symbol | Parameter IDs | Payload | Missing Paths | Unexpected Paths | Enum Mismatch | Cardinality | Status |",
1581
+ "|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|",
1582
+ ...facts.entries.map((entry) => `| ${entry.markdownModule} | ${entry.caseId ?? "—"} | ${entry.variantTestPoints.join(", ") || "—"} | ${entry.assertionTestPoints.join(", ") || "—"} | ${entry.crossCuttingTestPoints.join(", ") || "—"} | ${entry.mappedTestPoints.join(", ") || "—"} | ${entry.declaredScript ?? "—"} | ${entry.declaredPrimarySymbol ?? "—"} | ${entry.actualScripts.join(", ") || "—"} | ${entry.pytestSymbols.join(", ") || "—"} | ${entry.parameterIds.join(", ") || "—"} | ${entry.payloadAssessment.status} | ${entry.payloadAssessment.missingPaths.join(", ") || "—"} | ${entry.payloadAssessment.unexpectedPaths.join(", ") || "—"} | ${entry.payloadAssessment.enumMismatches.join(", ") || "—"} | ${entry.cardinality} | ${entry.status} |`),
1154
1583
  "",
1155
1584
  "## Findings",
1156
1585
  "",