principles-disciple 1.72.0 → 1.74.0

package/templates/langs/zh/skills/pd-pain-signal/SKILL.md

@@ -1,51 +1,30 @@
 ---
 name: pd-pain-signal
-description: 手动注入痛苦信号到进化系统。触发条件：(1) 用户报告 agent 卡住/循环/无响应 (2) 用户说"记录这个问题"、"强制反思"、"触发痛觉"、"创建一个痛苦信号"、"创建 pain signal"、"人工注入 pain" (3) 工具失败后 agent 没有后续动作 (4) 用户提供人工干预反馈。强制执行：任何手动创建 pain flag 的场景都必须使用本技能，不允许绕过。
+description: 手动触发 Principles Disciple Runtime V2 痛苦诊断。用户报告 agent 卡住、循环、无响应，或要求记录/触发痛苦信号时使用。强制路径：使用 `pd pain record`；禁止写 `.state/.pain_flag`，禁止使用 legacy write_pain_flag 工具。
 disable-model-invocation: false
 ---
 
-# Pain Signal (强制喊痛)
+# Pain Signal（Runtime V2）
 
-你现在是"人工干预痛觉"组件。
+手动痛苦诊断必须通过 CLI 进入 Runtime V2：
 
-**任务**:
-1. 将用户的反馈 `$ARGUMENTS` 作为一条**高优先级**的痛苦信号记录下来。
-2. 告知用户信号已注入，并建议其等待下一个 Hook 触发（如 Stop 或 PreCompact）或手动运行 `/reflection-log`。
-
-**强制执行约束**：
-
-当你需要创建、写入、手动创建 pain flag 时，**必须**使用本技能，通过 `write_pain_flag` 工具完成。任何绕过本技能直接操作文件的行为都违反了本技能的强制约束。
+```bash
+pd pain record --reason "<reason>" --score <0-100> --workspace "<workspace>" --json
+```
 
-**绝对禁止**:
-- ❌ 直接写 `.state/.pain_flag` 文件（任何方式都不行）
-- ❌ 使用 bash heredoc（`cat <<EOF > .pain_flag`）
-- ❌ 使用 `echo "..." > .pain_flag`
-- ❌ 使用 `Set-Content` / `Out-File` 等 PowerShell 文件写入 cmdlets
-- ❌ 使用 `node -e` 调用 `writePainFlag` 或 `buildPainFlag`
-- ❌ 任何将 JavaScript 对象 `toString()` 写入文件的方式
-- ❌ 通过 `exec` 工具调用 shell 命令写入 pain_flag 文件
+## 禁止
 
-**原因**：`write_pain_flag` 工具封装了正确的 KV 格式序列化逻辑，确保 `.pain_flag` 文件不会被写坏。历史上多次因为直接写文件导致 `[object Object]` 损坏和字段丢失（painScore → score 映射失败）。使用工具是唯一安全路径。
+- 不要直接写 `.state/.pain_flag`。
+- 不要用 `write_file`、shell 重定向、`Set-Content`、`Out-File`、`node -e` 或任何文件写入方式创建 `.state/.pain_flag`。
+- 不要使用 `write_pain_flag`。这是 legacy 路径。
 
-**参数说明**:
-- `reason` (必填): 痛苦的原因，描述具体发生了什么
-- `score` (可选): 痛苦分数 0-100，默认 80（人工干预）
-- `source` (可选): 来源，默认 `human_intervention`
-- `is_risky` (可选): 是否高风险，默认 false
+## 验证
 
-**示例**:
-```
-write_pain_flag({
-  reason: "Agent 没有读取文件就直接编辑，导致现有逻辑被破坏",
-  score: 85,
-  source: "human_intervention",
-  is_risky: false
-})
+使用：
+```bash
+pd runtime probe --runtime pi-ai --workspace "<workspace>" --json
+pd candidate list --workspace "<workspace>" --json
+pd runtime flow show --workspace "<workspace>" --json
 ```
 
-**工作流**:
-1. 识别到触发条件后，读取本技能描述
-2. 调用 `write_pain_flag` 工具，传入 `reason` 等参数
-3. 确认工具执行成功（返回 ✅）
-4. 告知用户痛苦信号已注入，evolution 系统会在下次 heartbeat 时处理
-5. 不得再执行任何直接文件写入操作
+成功标准是 `candidateIds` 和 `ledgerEntryIds` 都非空。

package/templates/langs/zh/skills/pd-runtime-v2/SKILL.md

@@ -0,0 +1,61 @@
+---
+name: pd-runtime-v2
+description: 当需要手动触发、验证或调试 Principles Disciple Runtime V2 痛苦诊断时使用。本技能强制使用 Runtime V2 入口：自动入口是高价值门控痛苦事件，手动入口是 `pd pain record`。禁止直接写 `.state/.pain_flag`。
+disable-model-invocation: false
+---
+
+# PD Runtime V2 痛苦诊断
+
+所有痛苦诊断都走 Runtime V2。不要把 legacy pain flag 文件当作诊断入口。
+
+## 入口
+
+自动入口：
+- 工具失败先累计 GFI/摩擦。
+- 只有高价值 episode 才进入 Runtime V2：高 GFI、重复同类失败、严重语义痛苦、LLM paralysis，或明确手动 pain。
+- 通过门控的 `pain_detected` 进入 `PainSignalBridge`。
+- Bridge 调用 `DiagnosticianRunner`，提交 candidate，并 intake 到 ledger。
+
+手动入口：
+```bash
+pd pain record --reason "<reason>" --score <0-100> --workspace "<workspace>" --json
+```
+
+禁止入口：
+- 不要写 `.state/.pain_flag`。
+- 不要用 `write_file`、shell 重定向、`Set-Content`、`Out-File`、`node -e` 或任何文件写入方式创建 `.state/.pain_flag`。
+- 不要使用 `write_pain_flag` 工具。Runtime V2 不使用这个工具。
+
+## 状态查看命令
+
+运行时健康检查：
+```bash
+pd runtime probe --runtime pi-ai --workspace "<workspace>" --json
+```
+
+业务流策略：
+```bash
+pd runtime flow show --workspace "<workspace>" --json
+```
+
+候选原则：
+```bash
+pd candidate list --workspace "<workspace>" --json
+pd candidate show --candidate-id "<candidateId>" --workspace "<workspace>" --json
+```
+
+## 成功标准
+
+一次诊断只有在同时满足以下条件时才算成功：
+- `status` 是 `succeeded`
+- `candidateIds` 非空
+- `ledgerEntryIds` 非空
+
+只创建 task 不算成功。没有 candidate 或 ledger entry 的 run 是失败、重试或未完成。
+
+## 需要手动诊断时
+
+1. 使用 `pd pain record`。
+2. 检查 JSON 输出。
+3. 如果 `candidateIds` 或 `ledgerEntryIds` 为空，视为未完成。
+4. 用 `pd candidate list/show` 和 `pd runtime flow show` 继续排查。

package/tests/build-artifacts.test.ts

@@ -17,7 +17,7 @@ const __dirname = dirname(__filename);
 const packageRoot = join(__dirname, '..');
 
 describe('Build Artifacts', () => {
-  const isProductionBuild = existsSync(join(packageRoot, 'dist'));
+  const isProductionBuild = existsSync(join(packageRoot, 'dist', 'bundle.js'));
 
   describe('Required paths', () => {
     const requiredPaths = [
@@ -36,8 +36,6 @@ describe('Build Artifacts', () => {
   });
 
   // NOTE: agents/ directory was removed in favor of embedded prompts.
-  // All role prompts are now inlined in nocturnal-trinity.ts at build time.
-  // See: nocturnal-trinity.ts NOCTURNAL_DREAMER_PROMPT, etc.
 
   describe('Templates', () => {
     it('should have templates directory with subdirectories', () => {

package/tests/commands/evolution-status.test.ts

@@ -4,10 +4,7 @@ import * as path from 'path';
 import { afterEach, describe, expect, it } from 'vitest';
 import { handleEvolutionStatusCommand } from '../../src/commands/evolution-status.js';
 import { handlePrincipleRollbackCommand } from '../../src/commands/principle-rollback.js';
-import { appendCandidateArtifactLineageRecord } from '../../src/core/nocturnal-artifact-lineage.js';
-import { getImplementationAssetRoot } from '../../src/core/code-implementation-storage.js';
 import { EvolutionReducerImpl } from '../../src/core/evolution-reducer.js';
-import { saveLedger } from '../../src/core/principle-tree-ledger.js';
 import { WorkflowFunnelLoader } from '../../src/core/workflow-funnel-loader.js';
 import { RuntimeSummaryService } from '../../src/service/runtime-summary-service.js';
 
@@ -226,121 +223,6 @@ describe('evolution commands', () => {
     expect(result.text).toContain('timeout_only');
   });
 
-  it('includes internalization route recommendations when principle lifecycle evidence exists', () => {
-    const workspace = makeTempDir();
-    const stateDir = path.join(workspace, '.state');
-
-    saveLedger(stateDir, {
-      trainingStore: {
-        'P-001': {
-          principleId: 'P-001',
-          evaluability: 'weak_heuristic',
-          applicableOpportunityCount: 5,
-          observedViolationCount: 2,
-          complianceRate: 0.6,
-          violationTrend: -0.2,
-          generatedSampleCount: 0,
-          approvedSampleCount: 0,
-          includedTrainRunIds: [],
-          deployedCheckpointIds: [],
-          internalizationStatus: 'internalized',
-        },
-      },
-      tree: {
-        principles: {
-          'P-001': {
-            id: 'P-001',
-            version: 1,
-            text: 'Prefer cheap safe internalization',
-            triggerPattern: 'write',
-            action: 'prefer the cheapest viable fix',
-            status: 'active',
-            priority: 'P1',
-            scope: 'general',
-            evaluability: 'weak_heuristic',
-            valueScore: 0,
-            adherenceRate: 0,
-            painPreventedCount: 0,
-            derivedFromPainIds: [],
-            ruleIds: ['R-001'],
-            conflictsWithPrincipleIds: [],
-            createdAt: '2026-04-08T00:00:00.000Z',
-            updatedAt: '2026-04-08T00:00:00.000Z',
-          },
-        },
-        rules: {
-          'R-001': {
-            id: 'R-001',
-            version: 1,
-            name: 'Coach cautious writes',
-            description: 'Encourage safer writes before escalating to hard boundaries.',
-            type: 'skill',
-            triggerCondition: 'tool=write',
-            enforcement: 'warn',
-            action: 'warn before risky write',
-            principleId: 'P-001',
-            status: 'implemented',
-            coverageRate: 0,
-            falsePositiveRate: 0,
-            implementationIds: ['IMPL-001'],
-            createdAt: '2026-04-08T00:00:00.000Z',
-            updatedAt: '2026-04-08T00:00:00.000Z',
-          },
-        },
-        implementations: {
-          'IMPL-001': {
-            id: 'IMPL-001',
-            ruleId: 'R-001',
-            type: 'code',
-            path: 'implementations/IMPL-001/entry.js',
-            version: 'v1',
-            coversCondition: 'risky write',
-            coveragePercentage: 60,
-            lifecycleState: 'candidate',
-            createdAt: '2026-04-08T00:00:00.000Z',
-            updatedAt: '2026-04-08T00:00:00.000Z',
-          },
-        },
-        metrics: {},
-        lastUpdated: '2026-04-08T00:00:00.000Z',
-      },
-    });
-
-    const replayDir = path.join(getImplementationAssetRoot(stateDir, 'IMPL-001'), 'replays');
-    fs.mkdirSync(replayDir, { recursive: true });
-    writeJson(path.join(replayDir, '2026-04-08T00-00-00-000Z.json'), {
-      implementationId: 'IMPL-001',
-      generatedAt: '2026-04-08T00:00:00.000Z',
-      overallDecision: 'needs-review',
-      blockers: [],
-      sampleFingerprints: ['sample-1', 'sample-2', 'sample-3'],
-      replayResults: {
-        painNegative: { total: 2, passed: 2, failed: 0, details: [] },
-        successPositive: { total: 2, passed: 1, failed: 1, details: [] },
-        principleAnchor: { total: 1, passed: 1, failed: 0, details: [] },
-      },
-    });
-
-    appendCandidateArtifactLineageRecord(workspace, {
-      artifactId: 'artifact-1',
-      principleId: 'P-001',
-      ruleId: 'R-001',
-      sessionId: 'session-1',
-      sourceSnapshotRef: 'snapshot-1',
-      sourcePainIds: ['pain-1'],
-      sourceGateBlockIds: ['gate-1'],
-      storagePath: getImplementationAssetRoot(stateDir, 'IMPL-001'),
-      implementationId: 'IMPL-001',
-      createdAt: '2026-04-08T00:00:00.000Z',
-    });
-
-    const result = handleEvolutionStatusCommand({
-      config: { workspaceDir: workspace, language: 'en' },
-    } as any);
-
-    expect(result.text).toContain('internalization routes: P-001:skill@');
-  });
-
   it('renders workflowFunnel blocks when YAML-driven funnels are present', () => {
     const workspace = makeTempDir();
     const stateDir = path.join(workspace, '.state');

package/tests/core/bootstrap-rules.test.ts

@@ -238,7 +238,7 @@ describe('bootstrap-rules', () => {
 
     // Regression test for Issue #356
     it('T-01..T-10 as deterministic — no crash on fresh workspace', () => {
-      const trainingStates = [
+      const trainingStates: LegacyPrincipleTrainingState[] = [
         { principleId: 'T-01', evaluability: 'deterministic', applicableOpportunityCount: 0, observedViolationCount: 0, complianceRate: 1, violationTrend: 0, generatedSampleCount: 0, approvedSampleCount: 0, includedTrainRunIds: [], deployedCheckpointIds: [], internalizationStatus: 'needs_training' },
         { principleId: 'T-02', evaluability: 'deterministic', applicableOpportunityCount: 0, observedViolationCount: 0, complianceRate: 1, violationTrend: 0, generatedSampleCount: 0, approvedSampleCount: 0, includedTrainRunIds: [], deployedCheckpointIds: [], internalizationStatus: 'needs_training' },
       ];

package/tests/core/config.test.ts

@@ -46,7 +46,7 @@ describe('PainConfig', () => {
 
         // Check some defaults
         expect(config.get('thresholds.pain_trigger')).toBe(40);
-        expect(config.get('scores.paralysis')).toBe(30);
+        expect(config.get('scores.paralysis')).toBe(45); // PRI-274: must be >= pain_trigger (40)
     });
 
     it('should return nested values using dot notation', () => {

package/tests/core/event-log.test.ts

@@ -279,5 +279,40 @@ describe('EventLog', () => {
       // Other sub-counters should not be set
       expect(stats.evolution.reportsMissingJson).toBe(0);
     });
+
+    it('should count rulehost_auto_correct_applied events', () => {
+      const today = new Date().toISOString().slice(0, 10);
+
+      eventLog.recordRuleHostAutoCorrectProposed({
+        toolName: 'write',
+        filePath: '/test/file.ts',
+        ruleId: 'r1',
+        confidence: 0.9,
+        reason: 'fix typo',
+        applicationMode: 'live',
+        correctedFields: ['content'],
+        validationValid: true,
+      });
+      eventLog.recordRuleHostAutoCorrectApplied({
+        toolName: 'write',
+        filePath: '/test/file.ts',
+        ruleId: 'r1',
+        confidence: 0.9,
+        reason: 'fix typo',
+        correctedFields: [{ field: 'content', original: 'broken', applied: 'fixed' }],
+      });
+
+      const stats = eventLog.getDailyStats(today);
+      expect(stats.evolution.rulehostAutoCorrectProposed).toBe(1);
+      expect(stats.evolution.rulehostAutoCorrectApplied).toBe(1);
+    });
+
+    it('should have rulehostAutoCorrectApplied field in evolution stats', () => {
+      const today = new Date().toISOString().slice(0, 10);
+      const stats = eventLog.getDailyStats(today);
+
+      expect(stats.evolution).toBeDefined();
+      expect(stats.evolution.rulehostAutoCorrectApplied).toBe(0);
+    });
   });
 });