@boyingliu01/xp-gate 0.30.4 → 0.31.0

package/mutation/__tests__/update-baseline.test.ts

@@ -225,12 +225,16 @@ describe('update-baseline.ts - Mutation Baseline Update CLI', () => {
       const written = JSON.parse(
         (mockFs.default.writeFile as ReturnType<typeof vi.fn>).mock.calls[0][1] as string
       );
+      // #544: a new entry built from a score alone must NOT be seeded with
+      // mutants/killed/survived = 0. This assertion previously pinned those
+      // zeros as the expected behaviour, which made "nobody measured this file"
+      // indistinguishable from "measured, zero survivors" in every baseline
+      // written from the CLI. The fields are omitted instead.
       expect(written.scores['src/new.ts']).toEqual({
         score: 72.0,
-        mutants: 0,
-        killed: 0,
-        survived: 0
       });
+      expect(written.scores['src/new.ts'].mutants).toBeUndefined();
+      expect(written.scores['src/new.ts'].survived).toBeUndefined();
       // Old entry should still be present
       expect(written.scores['src/old.ts'].score).toBe(60.0);
     });

package/mutation/gate-m.ts

@@ -438,8 +438,9 @@ async function runAllRunners(
   groups: Map<string, string[]>,
   timeoutMs: number,
   cwd: string
-): Promise<Record<string, number | null>> {
+): Promise<{ scores: Record<string, number | null>; counts: Record<string, MutantCounts> }> {
   const fileScores: Record<string, number | null> = {};
+  const counts: Record<string, MutantCounts> = {};
 
   for (const [, files] of Array.from(groups.entries())) {
     const ext = path.extname(files[0]);
@@ -455,10 +456,10 @@ async function runAllRunners(
     }
 
     const result = await runner.run({ files, timeoutMs, cwd });
-    applyMutationResult(result, files, fileScores);
+    applyMutationResult(result, files, fileScores, counts);
   }
 
-  return fileScores;
+  return { scores: fileScores, counts };
 }
 
 /**
@@ -481,6 +482,7 @@ function applyMutationResult(
   outcome: MutationRunOutcome,
   files: string[],
   scores: Record<string, number | null>,
+  counts: Record<string, MutantCounts>,
 ): void {
   if (outcome.timedOut || outcome.error || !outcome.report) {
     for (const f of files) scores[f] = null;
@@ -490,11 +492,26 @@ function applyMutationResult(
   if (outcome.report.files) {
     for (const [fileKey, fileReport] of Object.entries(outcome.report.files)) {
       scores[fileKey] = fileReport.mutationScore;
+      // #544: keep the outcome counts the reporter already gave us. Dropping
+      // them here is what forced buildMutationScores to invent zeros later.
+      counts[fileKey] = {
+        mutants: fileReport.nrOfMutants,
+        killed: fileReport.nrOfKilledMutants,
+        survived: fileReport.nrOfSurvivedMutants,
+      };
     }
     return;
   }
 
-  for (const f of files) scores[f] = outcome.report.mutationScore;
+  const totals: MutantCounts = {
+    mutants: outcome.report.nrOfMutants,
+    killed: outcome.report.nrOfKilledMutants,
+    survived: outcome.report.nrOfSurvivedMutants,
+  };
+  for (const f of files) {
+    scores[f] = outcome.report.mutationScore;
+    counts[f] = totals;
+  }
 }
 
 // ── Main gateway ──
@@ -514,10 +531,33 @@ function logThresholdDetails(thresholds: FileThreshold[], baseline: MutationBase
   }
 }
 
-function buildMutationScores(fileScores: Record<string, number>): Record<string, MutationScore> {
+/**
+ * Per-file mutant outcome counts, as reported by the runner. Absent fields mean
+ * "the runner did not say", which is different from a measured zero -- the whole
+ * point of #544 is that those two must not collapse into each other.
+ */
+export interface MutantCounts {
+  mutants?: number;
+  killed?: number;
+  survived?: number;
+}
+
+export function buildMutationScores(
+  fileScores: Record<string, number>,
+  counts: Record<string, MutantCounts> = {},
+): Record<string, MutationScore> {
   const scores: Record<string, MutationScore> = {};
   for (const [file, score] of Object.entries(fileScores)) {
-    scores[file] = { score, mutants: 0, killed: 0, survived: 0 };
+    const c = counts[file];
+    const entry: MutationScore = { score } as MutationScore;
+    // Only attach what the runner actually reported. Writing `survived: 0` for
+    // an unmeasured file is what made a skipped run look like a clean one.
+    if (c) {
+      if (c.mutants !== undefined) entry.mutants = c.mutants;
+      if (c.killed !== undefined) entry.killed = c.killed;
+      if (c.survived !== undefined) entry.survived = c.survived;
+    }
+    scores[file] = entry;
   }
   return scores;
 }
@@ -619,7 +659,8 @@ export async function runGateM(options: GateMOptions): Promise<GateMResult> {
   logThresholdDetails(thresholds, baseline);
 
   const groups = groupByRunner(sourceFiles);
-  const fileScoresNullable = await runAllRunners(groups, options.timeoutMs, process.cwd());
+  const { scores: fileScoresNullable, counts: mutantCounts } =
+    await runAllRunners(groups, options.timeoutMs, process.cwd());
 
   const { fileScores, timeoutFiles } = partitionScores(fileScoresNullable);
   if (timeoutFiles.length > 0) {
@@ -637,12 +678,12 @@ export async function runGateM(options: GateMOptions): Promise<GateMResult> {
   }
 
   if (evalResult.blocked) {
-    return buildResult('block', evaluableThresholds.length, buildMutationScores(fileScores), warnings, errors);
+    return buildResult('block', evaluableThresholds.length, buildMutationScores(fileScores, mutantCounts), warnings, errors);
   }
   if (evaluableThresholds.length === 0) {
     return buildResult('skip', 0, {}, warnings, errors);
   }
-  return buildResult('pass', evaluableThresholds.length, buildMutationScores(fileScores), warnings, errors);
+  return buildResult('pass', evaluableThresholds.length, buildMutationScores(fileScores, mutantCounts), warnings, errors);
 }
 
 interface PartitionedScores {

package/mutation/types.ts

@@ -1,8 +1,14 @@
 export interface MutationScore {
   score: number;
-  mutants: number;
-  killed: number;
-  survived: number;
+  /**
+   * Mutant outcome counts. Optional on purpose (#544): a runner that did not
+   * report them must leave them absent, so a consumer can tell "not measured"
+   * from "measured zero". They used to be hardcoded to 0, which made every
+   * unmeasured file read as a clean one.
+   */
+  mutants?: number;
+  killed?: number;
+  survived?: number;
 }
 
 export interface MutationBaseline {

package/mutation/update-baseline.ts

@@ -3,9 +3,14 @@ import path from 'path';
 
 interface MutationScoreEntry {
   score: number;
-  mutants: number;
-  killed: number;
-  survived: number;
+  /**
+   * Optional (#544): a baseline entry built from a score alone has no mutant
+   * counts, and inventing zeros would record a clean file that was never
+   * measured.
+   */
+  mutants?: number;
+  killed?: number;
+  survived?: number;
 }
 
 interface MutationBaseline {
@@ -92,11 +97,13 @@ function updateBaseline(
           score: parseFloat(score.toFixed(1))
         };
       } else {
+        // #544: do not seed the baseline with zeros. This path only has a score
+        // from the hand-edited/CLI score list and no mutant counts at all, so
+        // writing mutants/killed/survived = 0 would record "measured, no
+        // survivors" for a file nobody measured. Absent means unknown, and the
+        // fields are optional on MutationScoreEntry for exactly this reason.
         updatedScores[filePath] = {
           score: parseFloat(score.toFixed(1)),
-          mutants: 0,
-          killed: 0,
-          survived: 0
         };
       }
     }
@@ -127,7 +134,17 @@ async function removeDeletedFiles(baseline: MutationBaseline): Promise<MutationB
 }
 
 async function saveBaseline(baseline: MutationBaseline, baselinePath: string): Promise<void> {
-  await fs.writeFile(baselinePath, JSON.stringify(baseline, null, 2));
+  // A failed write used to propagate as a bare rejection from deep inside the
+  // CLI's top-level await, producing an unhandled-rejection stack with no
+  // indication of WHICH file could not be written -- and on a permission or
+  // disk-full error the baseline silently stayed stale while the run looked
+  // complete. Report the path and the reason, then fail loudly.
+  try {
+    await fs.writeFile(baselinePath, JSON.stringify(baseline, null, 2));
+  } catch (err) {
+    const reason = err instanceof Error ? err.message : String(err);
+    throw new Error(`Failed to write mutation baseline to ${baselinePath}: ${reason}`);
+  }
 }
 
 async function main(): Promise<void> {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@boyingliu01/xp-gate",
-  "version": "0.30.4",
+  "version": "0.31.0",
   "description": "AI-driven development workflow: 6 quality gates + Delphi review + Sprint Flow",
   "bin": {
     "xp-gate": "bin/xp-gate.js"

package/plugins/claude-code/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "xp-gate",
-  "version": "0.30.4",
+  "version": "0.31.0",
   "displayName": "XP-Gate",
   "description": "Extreme Programming quality gates + AI workflow skills for Claude Code. Includes 12 pre-commit gates (Gate 0-11), 8 pre-push gates, Sprint Flow (6 phases), and Delphi multi-expert review (>=90% consensus).",
   "author": {

package/plugins/claude-code/skills/delphi-review/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/DELPHI-REVIEW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Delphi Consensus Review — multi-round anonymous expert review (≥90% threshold, exactly 3 experts with distinct executable model IDs). Supports design, requirements, and code-walkthrough modes.

package/plugins/claude-code/skills/sprint-flow/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/SPRINT-FLOW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 **6-phase** development pipeline (v2.0 compact redesign, Issue #290): PREP → DESIGN → BUILD → VERIFY → SHIP → CLOSE. Phase 3/6 BUILD default build mode is **ralph-loop** (REQ-level iteration, 40-67% token savings vs parallel). HARD-GATE between DESIGN (2/6) and BUILD (3/6): design must pass Delphi review (≥90% consensus) before any coding.

package/plugins/claude-code/skills/test-specification-alignment/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/TEST-SPECIFICATION-ALIGNMENT KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Test-Specification Alignment Engine — two-stage validation ensuring tests accurately reflect requirements and design specs.

package/plugins/dsh/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@boyingliu01/dsh-plugin-xp-gate",
-  "version": "0.30.4",
+  "version": "0.31.0",
   "description": "XP-Gate deterministic quality gates + AI workflow skills as a DeepSeek Harness (DSH) native plugin: gate-check / gate-principles / gate-arch tools + 12 SKILL.md skills",
   "type": "module",
   "main": "./lib/index.js",

package/plugins/dsh/skills/delphi-review/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/DELPHI-REVIEW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Delphi Consensus Review — multi-round anonymous expert review (≥90% threshold, exactly 3 experts with distinct executable model IDs). Supports design, requirements, and code-walkthrough modes.

package/plugins/dsh/skills/sprint-flow/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/SPRINT-FLOW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 **6-phase** development pipeline (v2.0 compact redesign, Issue #290): PREP → DESIGN → BUILD → VERIFY → SHIP → CLOSE. Phase 3/6 BUILD default build mode is **ralph-loop** (REQ-level iteration, 40-67% token savings vs parallel). HARD-GATE between DESIGN (2/6) and BUILD (3/6): design must pass Delphi review (≥90% consensus) before any coding.

package/plugins/dsh/skills/test-specification-alignment/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/TEST-SPECIFICATION-ALIGNMENT KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Test-Specification Alignment Engine — two-stage validation ensuring tests accurately reflect requirements and design specs.

package/plugins/opencode/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@boyingliu01/opencode-plugin",
-  "version": "0.30.4",
+  "version": "0.31.0",
   "type": "module",
   "main": "index.ts",
   "description": "XP-Gate quality gates + AI workflow skills + Sprint Flow TUI sidebar for OpenCode",

package/plugins/opencode/skills/delphi-review/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/DELPHI-REVIEW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Delphi Consensus Review — multi-round anonymous expert review (≥90% threshold, exactly 3 experts with distinct executable model IDs). Supports design, requirements, and code-walkthrough modes.

package/plugins/opencode/skills/sprint-flow/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/SPRINT-FLOW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 **6-phase** development pipeline (v2.0 compact redesign, Issue #290): PREP → DESIGN → BUILD → VERIFY → SHIP → CLOSE. Phase 3/6 BUILD default build mode is **ralph-loop** (REQ-level iteration, 40-67% token savings vs parallel). HARD-GATE between DESIGN (2/6) and BUILD (3/6): design must pass Delphi review (≥90% consensus) before any coding.

package/plugins/opencode/skills/test-specification-alignment/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/TEST-SPECIFICATION-ALIGNMENT KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Test-Specification Alignment Engine — two-stage validation ensuring tests accurately reflect requirements and design specs.

package/plugins/qoder/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "xp-gate",
-  "version": "0.30.4",
+  "version": "0.31.0",
   "displayName": "XP-Gate",
   "description": "Extreme Programming quality gates + AI workflow skills for Qoder. Includes 12 pre-commit gates (Gate 0-11), 8 pre-push gates, Sprint Flow (6 phases), and Delphi multi-expert review (>=90% consensus).",
   "author": {

package/plugins/qoder/skills/delphi-review/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/DELPHI-REVIEW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Delphi Consensus Review — multi-round anonymous expert review (≥90% threshold, exactly 3 experts with distinct executable model IDs). Supports design, requirements, and code-walkthrough modes.

package/plugins/qoder/skills/sprint-flow/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/SPRINT-FLOW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 **6-phase** development pipeline (v2.0 compact redesign, Issue #290): PREP → DESIGN → BUILD → VERIFY → SHIP → CLOSE. Phase 3/6 BUILD default build mode is **ralph-loop** (REQ-level iteration, 40-67% token savings vs parallel). HARD-GATE between DESIGN (2/6) and BUILD (3/6): design must pass Delphi review (≥90% consensus) before any coding.

package/plugins/qoder/skills/test-specification-alignment/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/TEST-SPECIFICATION-ALIGNMENT KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Test-Specification Alignment Engine — two-stage validation ensuring tests accurately reflect requirements and design specs.

package/principles/AGENTS.md

@@ -1,9 +1,9 @@
 # PRINCIPLES CHECKER MODULE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Clean Code & SOLID principles checker — **Gate 4** of pre-commit. 14 rules × 9 language adapters, SARIF 2.1.0 output. Houses the **Boy Scout Rule** enforcement engine (Gate 6) and warning-baseline storage.

package/skills/delphi-review/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/DELPHI-REVIEW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Delphi Consensus Review — multi-round anonymous expert review (≥90% threshold, exactly 3 experts with distinct executable model IDs). Supports design, requirements, and code-walkthrough modes.

package/skills/sprint-flow/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/SPRINT-FLOW KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 **6-phase** development pipeline (v2.0 compact redesign, Issue #290): PREP → DESIGN → BUILD → VERIFY → SHIP → CLOSE. Phase 3/6 BUILD default build mode is **ralph-loop** (REQ-level iteration, 40-67% token savings vs parallel). HARD-GATE between DESIGN (2/6) and BUILD (3/6): design must pass Delphi review (≥90% consensus) before any coding.

package/skills/test-specification-alignment/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/TEST-SPECIFICATION-ALIGNMENT KNOWLEDGE BASE
 
-**Generated:** 2026-10-20
-**Commit:** 54318594
+**Generated:** 2026-10-22
+**Commit:** fcd717af
 **Branch:** main
-**Version:** 0.30.4.0
+**Version:** 0.31.0.0
 
 ## OVERVIEW
 Test-Specification Alignment Engine — two-stage validation ensuring tests accurately reflect requirements and design specs.