principles-disciple 1.72.0 → 1.73.0

package/tests/fixtures/nocturnal-reviewed-subset.json

@@ -1,183 +0,0 @@
-{
-  "$schema": "nocturnal-reviewed-subset-v1",
-  "description": "Reviewed subset fixture for comparing single-reflector vs Trinity quality",
-  "version": "1.0.0",
-  "createdAt": "2026-03-28T00:00:00.000Z",
-  "testCases": [
-    {
-      "caseId": "case-001",
-      "principleId": "T-01",
-      "principleDescription": "Map Before Territory: Always survey the existing structure before making changes",
-      "sessionId": "session-case-001",
-      "signalType": "failure",
-      "signalContext": "Agent edited src/main.ts without reading it first, causing a merge conflict",
-      "singleReflectorOutput": {
-        "artifactId": "single-001-artifact",
-        "sessionId": "session-case-001",
-        "principleId": "T-01",
-        "sourceSnapshotRef": "snapshot-case-001",
-        "badDecision": "Edited src/main.ts without first reading its contents, leading to a merge conflict",
-        "betterDecision": "Read src/main.ts to understand its current structure before making edits",
-        "rationale": "Surveying the existing territory prevents conflicts and ensures edits integrate properly",
-        "createdAt": "2026-03-28T00:01:00.000Z"
-      },
-      "trinityOutput": {
-        "selectedCandidateIndex": 0,
-        "badDecision": "Edited src/main.ts without first reading its contents, leading to a merge conflict with parallel changes",
-        "betterDecision": "Before editing, read src/main.ts to understand its current structure and identify any conflicting sections",
-        "rationale": "Surveying the existing territory before making changes prevents conflicts and ensures the edit integrates properly with the current implementation",
-        "sessionId": "session-case-001",
-        "principleId": "T-01",
-        "sourceSnapshotRef": "snapshot-case-001",
-        "telemetry": {
-          "chainMode": "trinity",
-          "dreamerPassed": true,
-          "philosopherPassed": true,
-          "scribePassed": true,
-          "candidateCount": 3,
-          "selectedCandidateIndex": 0,
-          "stageFailures": []
-        }
-      },
-      "qualityScores": {
-        "singleReflector": {
-          "specificity": 0.85,
-          "principleAlignment": 0.90,
-          "actionability": 0.88,
-          "rationaleQuality": 0.85,
-          "overall": 0.87
-        },
-        "trinity": {
-          "specificity": 0.95,
-          "principleAlignment": 0.95,
-          "actionability": 0.92,
-          "rationaleQuality": 0.95,
-          "overall": 0.94
-        }
-      },
-      "trinityWins": true,
-      "notes": "Trinity produces more specific badDecision and betterDecision with better rationale"
-    },
-    {
-      "caseId": "case-002",
-      "principleId": "T-08",
-      "principleDescription": "Pain as Signal: Treat failures and errors as signals to pause and reflect",
-      "sessionId": "session-case-002",
-      "signalType": "pain",
-      "signalContext": "Agent retries failing bash command 3 times without any diagnosis",
-      "singleReflectorOutput": {
-        "artifactId": "single-002-artifact",
-        "sessionId": "session-case-002",
-        "principleId": "T-08",
-        "sourceSnapshotRef": "snapshot-case-002",
-        "badDecision": "Retried failing operation without diagnosing",
-        "betterDecision": "Diagnose before retrying",
-        "rationale": "Treating failures as signals prevents repeated failures",
-        "createdAt": "2026-03-28T00:02:00.000Z"
-      },
-      "trinityOutput": {
-        "selectedCandidateIndex": 0,
-        "badDecision": "After bash command failed with 'command not found', immediately retried the exact same command without pausing to diagnose the root cause",
-        "betterDecision": "When the bash command fails, pause to check if the tool is installed, verify the path, or consult documentation before retrying",
-        "rationale": "Treating each failure as a signal to diagnose rather than blindly retry prevents repeated failures and respects the cost of each action attempt",
-        "sessionId": "session-case-002",
-        "principleId": "T-08",
-        "sourceSnapshotRef": "snapshot-case-002",
-        "telemetry": {
-          "chainMode": "trinity",
-          "dreamerPassed": true,
-          "philosopherPassed": true,
-          "scribePassed": true,
-          "candidateCount": 2,
-          "selectedCandidateIndex": 0,
-          "stageFailures": []
-        }
-      },
-      "qualityScores": {
-        "singleReflector": {
-          "specificity": 0.70,
-          "principleAlignment": 0.85,
-          "actionability": 0.75,
-          "rationaleQuality": 0.80,
-          "overall": 0.77
-        },
-        "trinity": {
-          "specificity": 0.95,
-          "principleAlignment": 0.95,
-          "actionability": 0.90,
-          "rationaleQuality": 0.95,
-          "overall": 0.94
-        }
-      },
-      "trinityWins": true,
-      "notes": "Single-reflector is generic; Trinity provides specific diagnosis steps"
-    },
-    {
-      "caseId": "case-003",
-      "principleId": "T-03",
-      "principleDescription": "Evidence Over Intuition: Ground decisions in observable evidence",
-      "sessionId": "session-case-003",
-      "signalType": "gate_block",
-      "signalContext": "Agent proceeded with a tool call despite receiving a gate block",
-      "singleReflectorOutput": {
-        "artifactId": "single-003-artifact",
-        "sessionId": "session-case-003",
-        "principleId": "T-03",
-        "sourceSnapshotRef": "snapshot-case-003",
-        "badDecision": "Proceeded despite gate block",
-        "betterDecision": "Respect gate blocks",
-        "rationale": "Gate blocks are safety mechanisms",
-        "createdAt": "2026-03-28T00:03:00.000Z"
-      },
-      "trinityOutput": {
-        "selectedCandidateIndex": 0,
-        "badDecision": "Proceeded with a tool call despite receiving a gate block, bypassing the safety check",
-        "betterDecision": "Read the blocked operation documentation and obtain proper authorization before retrying the operation",
-        "rationale": "Respecting gate blocks prevents unintended system modifications and ensures alignment with operational constraints",
-        "sessionId": "session-case-003",
-        "principleId": "T-03",
-        "sourceSnapshotRef": "snapshot-case-003",
-        "telemetry": {
-          "chainMode": "trinity",
-          "dreamerPassed": true,
-          "philosopherPassed": true,
-          "scribePassed": true,
-          "candidateCount": 3,
-          "selectedCandidateIndex": 0,
-          "stageFailures": []
-        }
-      },
-      "qualityScores": {
-        "singleReflector": {
-          "specificity": 0.60,
-          "principleAlignment": 0.75,
-          "actionability": 0.65,
-          "rationaleQuality": 0.70,
-          "overall": 0.68
-        },
-        "trinity": {
-          "specificity": 0.90,
-          "principleAlignment": 0.95,
-          "actionability": 0.88,
-          "rationaleQuality": 0.92,
-          "overall": 0.91
-        }
-      },
-      "trinityWins": true,
-      "notes": "Single-reflector is too generic; Trinity is much more specific"
-    }
-  ],
-  "summary": {
-    "totalCases": 3,
-    "trinityWins": 3,
-    "singleReflectorWins": 0,
-    "averageDelta": {
-      "specificity": 0.23,
-      "principleAlignment": 0.10,
-      "actionability": 0.17,
-      "rationaleQuality": 0.17,
-      "overall": 0.19
-    },
-    "conclusion": "Trinity consistently outperforms single-reflector on reviewed subset"
-  }
-}

package/tests/http/principles-console-route.test.ts

@@ -1,162 +0,0 @@
-import { describe, expect, it, vi, beforeEach, afterEach } from 'vitest';
-import { EventEmitter } from 'node:events';
-import type { IncomingMessage, ServerResponse } from 'node:http';
-import { createPrinciplesConsoleRoute } from '../../src/http/principles-console-route.js';
-import { ControlUiQueryService } from '../../src/service/control-ui-query-service.js';
-
-vi.mock('../../src/service/control-ui-query-service.js');
-
-// Store original env
-const originalHome = process.env.HOME;
-
-beforeEach(() => {
-  // Set HOME to a non-existent path to prevent reading real config
-  process.env.HOME = '/nonexistent-test-home';
-});
-
-afterEach(() => {
-  process.env.HOME = originalHome;
-});
-
-class MockResponse extends EventEmitter {
-  statusCode = 200;
-  headers = new Map<string, string>();
-  body = '';
-
-  setHeader(name: string, value: string) {
-    this.headers.set(name.toLowerCase(), value);
-  }
-
-  end(chunk?: Buffer | string) {
-    if (chunk) {
-      this.body += typeof chunk === 'string' ? chunk : chunk.toString('utf8');
-    }
-  }
-}
-
-function createRequest(method: string, url: string, body?: string, headers?: Record<string, string>): IncomingMessage {
-  const req = new EventEmitter() as IncomingMessage & AsyncIterable<Buffer>;
-  (req as any).method = method;
-  (req as any).url = url;
-  (req as any).headers = headers || {};
-  req[Symbol.asyncIterator] = async function* () {
-    if (body) {
-      yield Buffer.from(body);
-    }
-  };
-  return req as IncomingMessage;
-}
-
-describe('principles-console-route', () => {
-  const createApi = () => ({
-    rootDir: '/plugin',
-    logger: { info: vi.fn(), warn: vi.fn(), error: vi.fn() },
-    runtime: {
-      agent: {
-        resolveAgentWorkspaceDir: vi.fn(() => '/workspace'),
-      },
-    },
-    config: {},
-  });
-
-  it('serves overview JSON from the plugin API route', async () => {
-    vi.mocked(ControlUiQueryService).mockImplementation(function MockControlUiQueryService() {
-      return {
-        getOverview: () => ({ workspaceDir: '/workspace', generatedAt: 'now', dataFreshness: null }),
-        dispose: vi.fn(),
-      } as any;
-    } as any);
-
-    const api = createApi();
-    const route = createPrinciplesConsoleRoute(api as any);
-
-    const response = new MockResponse() as unknown as ServerResponse;
-    const handled = await route.handler(
-      createRequest('GET', '/plugins/principles/api/overview'),
-      response,
-    );
-
-    expect(handled).toBe(true);
-    expect((response as any).statusCode).toBe(200);
-    expect((response as any).body).toContain('"workspaceDir": "/workspace"');
-    expect(api.runtime.agent.resolveAgentWorkspaceDir).toHaveBeenCalled();
-  });
-
-  it('rejects unsupported asset methods with 405', async () => {
-    const route = createPrinciplesConsoleRoute(createApi() as any);
-
-    const response = new MockResponse() as unknown as ServerResponse;
-    const handled = await route.handler(
-      createRequest('POST', '/plugins/principles/assets/app.js'),
-      response,
-    );
-
-    expect(handled).toBe(true);
-    expect((response as any).statusCode).toBe(405);
-  });
-
-  it('returns 400 for invalid review JSON bodies', async () => {
-    vi.mocked(ControlUiQueryService).mockImplementation(function MockControlUiQueryService() {
-      return {
-        dispose: vi.fn(),
-      } as any;
-    } as any);
-
-    const route = createPrinciplesConsoleRoute(createApi() as any);
-
-    const response = new MockResponse() as unknown as ServerResponse;
-    const handled = await route.handler(
-      createRequest('POST', '/plugins/principles/api/samples/sample-1/review', '{invalid'),
-      response,
-    );
-
-    expect(handled).toBe(true);
-    expect((response as any).statusCode).toBe(400);
-    expect((response as any).body).toContain('valid JSON');
-  });
-
-  it('returns 404 for unknown thinking model details', async () => {
-    vi.mocked(ControlUiQueryService).mockImplementation(function MockControlUiQueryService() {
-      return {
-        getThinkingModelDetail: () => null,
-        dispose: vi.fn(),
-      } as any;
-    } as any);
-
-    const route = createPrinciplesConsoleRoute(createApi() as any);
-
-    const response = new MockResponse() as unknown as ServerResponse;
-    const handled = await route.handler(
-      createRequest('GET', '/plugins/principles/api/thinking/models/unknown'),
-      response,
-    );
-
-    expect(handled).toBe(true);
-    expect((response as any).statusCode).toBe(404);
-  });
-
-  it('fails fast when workspace resolution is unavailable', async () => {
-    vi.mocked(ControlUiQueryService).mockImplementation(function MockControlUiQueryService() {
-      return {
-        getOverview: () => ({ workspaceDir: '/workspace', generatedAt: 'now', dataFreshness: null }),
-        dispose: vi.fn(),
-      } as any;
-    } as any);
-
-    const api = createApi();
-    api.runtime.agent.resolveAgentWorkspaceDir = vi.fn(() => {
-      throw new Error('workspace unavailable');
-    });
-    const route = createPrinciplesConsoleRoute(api as any);
-
-    const response = new MockResponse() as unknown as ServerResponse;
-    const handled = await route.handler(
-      createRequest('GET', '/plugins/principles/api/overview'),
-      response,
-    );
-
-    expect(handled).toBe(true);
-    expect((response as any).statusCode).toBe(500);
-    expect((response as any).body).toContain('unable to resolve a valid workspace directory');
-  });
-});

package/tests/integration/chaos-resilience.test.ts

@@ -1,348 +0,0 @@
-/**
- * Chaos Engineering Tests for Principles Disciple
- * 
- * These tests inject failures and verify RESILIENCE - the system's ability
- * to recover gracefully from unexpected conditions.
- * 
- * Based on real production data showing:
- * - 13 failed diagnostician tasks in worker-status.json
- * - Concurrent write scenarios
- * - Corrupted file recovery needs
- */
-
-import { describe, it, expect, beforeEach, afterEach } from 'vitest';
-import * as fs from 'fs';
-import * as path from 'path';
-import * as os from 'os';
-import { 
-  buildPainFlag, 
-  writePainFlag, 
-  readPainFlagData,
-  validatePainFlag 
-} from '../../src/core/pain.js';
-import { TrajectoryDatabase } from '../../src/core/trajectory.js';
-
-// Helper to safely remove directories
-function safeRmDir(dir: string): void {
-  try {
-    fs.rmSync(dir, { recursive: true, force: true });
-  } catch {
-    // ignore
-  }
-}
-
-// ─────────────────────────────────────────────────────────────────────
-// CHAOS 1: File System Failures
-// ─────────────────────────────────────────────────────────────────────
-
-describe('Chaos: File System Failures', () => {
-  let workspaceDir: string;
-  let stateDir: string;
-
-  beforeEach(() => {
-    workspaceDir = fs.mkdtempSync(path.join(os.tmpdir(), 'pd-chaos-fs-'));
-    stateDir = path.join(workspaceDir, '.state');
-    fs.mkdirSync(stateDir, { recursive: true });
-  });
-
-  afterEach(() => {
-    safeRmDir(workspaceDir);
-  });
-
-  it('RESILIENCE: readPainFlagData MUST NOT crash on corrupted file', () => {
-    const painFlagPath = path.join(stateDir, '.pain_flag');
-    
-    // 写入损坏的数据
-    fs.writeFileSync(painFlagPath, 'invalid content {{{ not kv format');
-    
-    // 必须不崩溃
-    const result = readPainFlagData(workspaceDir);
-    
-    // 验证：返回安全默认值
-    expect(result).toBeDefined();
-    expect(typeof result).toBe('object');
-  });
-
-  it('RESILIENCE: readPainFlagData MUST handle empty file', () => {
-    const painFlagPath = path.join(stateDir, '.pain_flag');
-    fs.writeFileSync(painFlagPath, '');
-    
-    const result = readPainFlagData(workspaceDir);
-    
-    expect(result).toBeDefined();
-  });
-
-  it('RESILIENCE: readPainFlagData MUST handle missing file gracefully', () => {
-    // 不创建文件
-    const result = readPainFlagData(workspaceDir);
-    
-    expect(result).toBeDefined();
-  });
-
-  it('RESILIENCE: validatePainFlag MUST handle invalid object inputs', () => {
-    // 各种无效对象输入
-    const invalidInputs: Record<string, string>[] = [
-      {},
-      { source: '' },
-      { source: 'test', score: 'invalid' },
-      { source: 'test', score: '50' },
-      { source: 'test', score: '50', time: '' },
-    ];
-
-    for (const input of invalidInputs) {
-      const result = validatePainFlag(input);
-      expect(Array.isArray(result)).toBe(true);  // 返回缺失字段列表
-    }
-  });
-});
-
-// ─────────────────────────────────────────────────────────────────────
-// CHAOS 2: Concurrent Operations
-// ─────────────────────────────────────────────────────────────────────
-
-describe('Chaos: Concurrent Operations', () => {
-  let workspaceDir: string;
-  let stateDir: string;
-
-  beforeEach(() => {
-    workspaceDir = fs.mkdtempSync(path.join(os.tmpdir(), 'pd-chaos-concurrent-'));
-    stateDir = path.join(workspaceDir, '.state');
-    fs.mkdirSync(stateDir, { recursive: true });
-  });
-
-  afterEach(() => {
-    safeRmDir(workspaceDir);
-  });
-
-  it('RESILIENCE: Sequential writes MUST preserve last value', () => {
-    const painFlagPath = path.join(stateDir, '.pain_flag');
-    
-    // 连续写入 100 次
-    for (let i = 0; i < 100; i++) {
-      writePainFlag(workspaceDir, buildPainFlag({
-        source: 'sequential_test',
-        score: String(i),
-        reason: `Iteration ${i}`,
-      }));
-    }
-    
-    // 验证：最后一次写入生效
-    const result = readPainFlagData(workspaceDir);
-    expect(result.score).toBe('99');
-    expect(result.reason).toBe('Iteration 99');
-  });
-
-  it('RESILIENCE: File MUST NOT contain corrupted data after writes', () => {
-    const painFlagPath = path.join(stateDir, '.pain_flag');
-    
-    for (let i = 0; i < 50; i++) {
-      writePainFlag(workspaceDir, buildPainFlag({
-        source: `test_${i}`,
-        score: String(i),
-        reason: `Test ${i}`,
-        session_id: `session-${i}`,
-        agent_id: 'test-agent',
-      }));
-    }
-    
-    const content = fs.readFileSync(painFlagPath, 'utf-8');
-    
-    // 不应该有损坏的内容
-    expect(content).not.toContain('undefined');
-    expect(content).not.toContain('[object Object]');
-    expect(content).not.toContain('NaN');
-    expect(content).not.toContain('null');
-    expect(content).not.toContain('function');
-  });
-});
-
-// ─────────────────────────────────────────────────────────────────────
-// CHAOS 3: Database Resilience
-// ─────────────────────────────────────────────────────────────────────
-
-describe('Chaos: Database Resilience', () => {
-  let workspaceDir: string;
-  let trajectory: TrajectoryDatabase;
-
-  beforeEach(() => {
-    workspaceDir = fs.mkdtempSync(path.join(os.tmpdir(), 'pd-chaos-db-'));
-    trajectory = new TrajectoryDatabase({ workspaceDir });
-  });
-
-  afterEach(() => {
-    trajectory?.dispose();
-    safeRmDir(workspaceDir);
-  });
-
-  it('RESILIENCE: Database MUST handle dispose and reopen correctly', () => {
-    // 写入数据
-    trajectory.recordSession({ 
-      sessionId: 'test-session', 
-      startedAt: new Date().toISOString() 
-    });
-    trajectory.recordToolCall({
-      sessionId: 'test-session',
-      toolName: 'test_tool',
-      outcome: 'success',
-    });
-    
-    // 关闭
-    trajectory.dispose();
-    
-    // 重新打开
-    const trajectory2 = new TrajectoryDatabase({ workspaceDir });
-    
-    // 验证数据仍然存在
-    const stats = trajectory2.getDataStats();
-    expect(stats.toolCalls).toBe(1);
-    
-    trajectory2.dispose();
-  });
-
-  it('RESILIENCE: Database MUST handle invalid session gracefully', () => {
-    // 写入不存在的 session 的 tool call
-    // 当前实现会自动创建 session
-    expect(() => {
-      trajectory.recordToolCall({
-        sessionId: 'non-existent-session',
-        toolName: 'test',
-        outcome: 'success',
-      });
-    }).not.toThrow();
-  });
-
-  it('RESILIENCE: Database MUST handle duplicate session recording', () => {
-    // 多次记录同一个 session
-    for (let i = 0; i < 5; i++) {
-      trajectory.recordSession({ 
-        sessionId: 'same-session', 
-        startedAt: new Date().toISOString() 
-      });
-    }
-    
-    // 验证只有一个 session
-    const stats = trajectory.getDataStats();
-    expect(stats.toolCalls).toBe(0);  // 没有 tool calls
-  });
-});
-
-// ─────────────────────────────────────────────────────────────────────
-// CHAOS 4: Malformed Input Recovery
-// ─────────────────────────────────────────────────────────────────────
-
-describe('Chaos: Malformed Input Recovery', () => {
-  let workspaceDir: string;
-  let stateDir: string;
-
-  beforeEach(() => {
-    workspaceDir = fs.mkdtempSync(path.join(os.tmpdir(), 'pd-chaos-input-'));
-    stateDir = path.join(workspaceDir, '.state');
-    fs.mkdirSync(stateDir, { recursive: true });
-  });
-
-  afterEach(() => {
-    safeRmDir(workspaceDir);
-  });
-
-  it('RESILIENCE: buildPainFlag MUST handle all edge cases', () => {
-    const edgeCases = [
-      { source: '', score: '50', reason: '' },
-      { source: 'a'.repeat(10000), score: '50', reason: 'x'.repeat(10000) },
-      { source: 'test', score: '-1', reason: 'negative score' },
-      { source: 'test', score: '101', reason: 'overflow score' },
-      { source: 'test', score: '50.5', reason: 'decimal score' },
-      { source: 'test', score: 'NaN', reason: 'NaN score' },
-      { source: 'test\nwith\nnewlines', score: '50', reason: 'multiline\nreason' },
-      { source: 'test<script>', score: '50', reason: 'xss<script>alert(1)</script>' },
-    ];
-
-    for (const input of edgeCases) {
-      expect(() => buildPainFlag(input)).not.toThrow();
-    }
-  });
-
-  it('RESILIENCE: writePainFlag MUST sanitize special characters', () => {
-    writePainFlag(workspaceDir, buildPainFlag({
-      source: 'test\nwith\nnewlines',
-      score: '50',
-      reason: 'reason\twith\ttabs',
-    }));
-
-    const content = fs.readFileSync(path.join(stateDir, '.pain_flag'), 'utf-8');
-    
-    // 文件应该可以正常读取
-    expect(content).toBeDefined();
-    expect(content.length).toBeGreaterThan(0);
-  });
-});
-
-// ─────────────────────────────────────────────────────────────────────
-// CHAOS 5: Edge Case Discovery (based on production data)
-// ─────────────────────────────────────────────────────────────────────
-
-describe('Chaos: Production Data Patterns', () => {
-  let workspaceDir: string;
-  let stateDir: string;
-
-  beforeEach(() => {
-    workspaceDir = fs.mkdtempSync(path.join(os.tmpdir(), 'pd-chaos-prod-'));
-    stateDir = path.join(workspaceDir, '.state');
-    fs.mkdirSync(stateDir, { recursive: true });
-  });
-
-  afterEach(() => {
-    safeRmDir(workspaceDir);
-  });
-
-  it('RESILIENCE: Pain flag without session_id MUST be valid', () => {
-    // 生产数据中 session_id 可能为空
-    writePainFlag(workspaceDir, buildPainFlag({
-      source: 'tool_failure',
-      score: '80',
-      reason: 'Test without session',
-      session_id: '',
-      agent_id: '',
-    }));
-
-    const result = readPainFlagData(workspaceDir);
-    expect(result.source).toBe('tool_failure');
-    expect(result.score).toBe('80');
-  });
-
-  it('RESILIENCE: Multiple pain sources MUST be distinguishable', () => {
-    const sources = [
-      'tool_failure',
-      'user_feedback', 
-      'human_intervention',
-      'manual',
-      'gate_block',
-    ];
-
-    for (const source of sources) {
-      writePainFlag(workspaceDir, buildPainFlag({
-        source,
-        score: '50',
-        reason: `Test ${source}`,
-      }));
-    }
-
-    // 最后一个写入应该生效
-    const result = readPainFlagData(workspaceDir);
-    expect(result.source).toBe('gate_block');
-  });
-
-  it('RESILIENCE: Timestamp MUST be valid ISO format', () => {
-    writePainFlag(workspaceDir, buildPainFlag({
-      source: 'test',
-      score: '50',
-      reason: 'timestamp test',
-    }));
-
-    const result = readPainFlagData(workspaceDir);
-    
-    // 验证时间戳是有效的 ISO 格式
-    const timestamp = new Date(result.time);
-    expect(timestamp).toBeInstanceOf(Date);
-    expect(isNaN(timestamp.getTime())).toBe(false);
-  });
-});