tribunal-kit 9.2.1 → 9.2.4

package/.agent/history/snapshots/test__unit__token_budget_broker.test.js.json

@@ -0,0 +1,10 @@
+{
+  "file": "test/unit/token_budget_broker.test.js",
+  "riskScore": "Low",
+  "blastRadius": 0,
+  "imports": {
+    "../../.agent/scripts/token_budget_broker": []
+  },
+  "dependents": [],
+  "content": "/**\n * token_budget_broker.test.js - Tests for Token Budget Broker\n */\n\nconst {\n  getTokenBudget,\n  TIER_TOKEN_LIMITS: _TIER_TOKEN_LIMITS,\n  MODEL_CONTEXT_LIMITS,\n} = require('../../.agent/scripts/token_budget_broker');\n\ndescribe('token_budget_broker.js', () => {\n  describe('getTokenBudget()', () => {\n    // Test backward compatibility with original signature\n    describe('Backward Compatibility', () => {\n      test('returns correct values for Tier 0', () => {\n        const budget = getTokenBudget(0);\n        expect(budget.tier).toBe(0);\n        expect(budget.maxTokens).toBe(0);\n        expect(budget.includeFullRepo).toBe(false);\n        expect(budget.maxSkills).toBe(0);\n        expect(budget.maxReviewers).toBe(0);\n      });\n\n      test('returns correct values for Tier 1', () => {\n        const budget = getTokenBudget(1);\n        expect(budget.tier).toBe(1);\n        expect(budget.maxTokens).toBe(2000);\n        expect(budget.includeFullRepo).toBe(false);\n        expect(budget.maxSkills).toBe(1);\n        expect(budget.maxReviewers).toBe(1);\n      });\n\n      test('returns correct values for Tier 2', () => {\n        const budget = getTokenBudget(2);\n        expect(budget.tier).toBe(2);\n        expect(budget.maxTokens).toBe(8000);\n        expect(budget.includeFullRepo).toBe(true);\n        expect(budget.maxSkills).toBe(3);\n        expect(budget.maxReviewers).toBe(2);\n      });\n\n      test('returns correct values for Tier 3', () => {\n        const budget = getTokenBudget(3);\n        expect(budget.tier).toBe(3);\n        expect(budget.maxTokens).toBe(32000);\n        expect(budget.includeFullRepo).toBe(true);\n        expect(budget.maxSkills).toBe(10);\n        expect(budget.maxReviewers).toBe(8);\n      });\n\n      test('normalizes invalid tiers', () => {\n        expect(getTokenBudget(-1).tier).toBe(0);\n        expect(getTokenBudget(4).tier).toBe(3);\n        expect(getTokenBudget(1.5).tier).toBe(2); // 1.5 rounds to 2\n      });\n    });\n\n    // Test new dynamic functionality\n    describe('Dynamic Budget Calculation', () => {\n      test('adjusts budget based on model context limit', () => {\n        // Claude Sonnet 5: 200k tokens\n        // Tier 1 ratio: 0.02\n        // Expected: 200000 * 0.02 = 4000, but capped at TIER_TOKEN_LIMITS[1] = 2000\n        const budget = getTokenBudget(1, 'claude-sonnet-5', 0);\n        expect(budget.maxTokens).toBe(2000); // Should be capped at static limit\n        expect(budget.modelLimit).toBe(200000);\n        expect(budget.availableForContext).toBe(190000); // 200000 - 0 - 10000 reserve\n        expect(budget.dynamicMaxTokens).toBe(3800); // 190000 * 0.02\n      });\n\n      test('uses lower limit for smaller models', () => {\n        // GPT-4o mini: 128k tokens\n        // Tier 2 ratio: 0.06\n        // Expected: 128000 * 0.06 = 7680, but capped at TIER_TOKEN_LIMITS[2] = 8000\n        const budget = getTokenBudget(2, 'gpt-4o-mini', 0);\n        expect(budget.maxTokens).toBe(7680); // Not capped since 7680 < 8000\n        expect(budget.modelLimit).toBe(128000);\n      });\n\n      test('accounts for conversation history tokens', () => {\n        // With 50k history tokens, less available for context\n        const budget = getTokenBudget(2, 'claude-sonnet-5', 50000);\n        // Available: 200000 - 50000 - 10000 = 140000\n        // Dynamic: 140000 * 0.06 = 8400\n        // Capped at: 8000\n        expect(budget.maxTokens).toBe(8000);\n        expect(budget.availableForContext).toBe(140000);\n        expect(budget.dynamicMaxTokens).toBe(8400);\n      });\n\n      test('handles zero available context', () => {\n        // If history tokens exceed available space\n        const budget = getTokenBudget(2, 'claude-sonnet-5', 190000);\n        // Available: 200000 - 190000 - 10000 = 0\n        expect(budget.maxTokens).toBe(0);\n        expect(budget.availableForContext).toBe(0);\n        expect(budget.dynamicMaxTokens).toBe(0);\n      });\n\n      test('respects model context limits', () => {\n        // Test with Gemini 2.5 Pro: 1M tokens\n        const budget = getTokenBudget(3, 'gemini-2.5-pro', 0);\n        // Available: 1000000 - 0 - 10000 = 990000\n        // Dynamic: 990000 * 0.25 = 247500\n        // But capped at TIER_TOKEN_LIMITS[3] = 32000\n        expect(budget.maxTokens).toBe(32000);\n        expect(budget.modelLimit).toBe(1000000);\n      });\n\n      test('uses default model when unknown model provided', () => {\n        const budget = getTokenBudget(1, 'unknown-model', 0);\n        expect(budget.modelLimit).toBe(MODEL_CONTEXT_LIMITS['claude-sonnet-5']);\n      });\n    });\n\n    // Test density scoring\n    describe('Context Density Scoring', () => {\n      test('applies density factor when skill metrics provided', () => {\n        const skillMetrics = { densityScore: 1.0 }; // Maximum density\n        const budget = getTokenBudget(1, 'claude-sonnet-5', 0, skillMetrics);\n        // Base budget for Tier 1: 2000\n        // Density factor for score 1.0: 0.5 + (1.0 * 1.5) = 2.0\n        // Expected: 2000 * 2.0 = 4000, but capped at 2000\n        expect(budget.maxTokens).toBe(2000); // Still capped\n        expect(budget.densityFactor).toBe(2.0);\n      });\n\n      test('applies density factor within limits', () => {\n        const skillMetrics = { densityScore: 0.5 }; // Medium density\n        const budget = getTokenBudget(2, 'claude-sonnet-5', 0, skillMetrics);\n        // Base budget for Tier 2: 8000\n        // Density factor for score 0.5: 0.5 + (0.5 * 1.5) = 1.25\n        // Expected: 8000 * 1.25 = 10000\n        expect(budget.maxTokens).toBe(10000);\n        expect(budget.densityFactor).toBe(1.25);\n      });\n\n      test('combines multiple density metrics', () => {\n        const skillMetrics = {\n          densityScore: 0.5,\n          ruleDensity: 0.5,\n          codeExampleDensity: 0.5,\n        };\n        const budget = getTokenBudget(1, 'claude-sonnet-5', 0, skillMetrics);\n        // Should have applied all three metrics\n        expect(budget.maxTokens).toBeGreaterThan(2000);\n        expect(budget.densityFactor).toBeGreaterThan(1.0);\n      });\n\n      test('clamps density factor to reasonable range', () => {\n        // Test with extreme values\n        const skillMetrics = { densityScore: 2.0 }; // Way above max\n        const budget = getTokenBudget(1, 'claude-sonnet-5', 0, skillMetrics);\n        expect(budget.densityFactor).toBeLessThanOrEqual(2.0);\n\n        const skillMetrics2 = { densityScore: -1.0 }; // Below min\n        const budget2 = getTokenBudget(1, 'claude-sonnet-5', 0, skillMetrics2);\n        expect(budget2.densityFactor).toBeGreaterThanOrEqual(0.5);\n      });\n    });\n  });\n});\n"
+}

package/.agent/history/snapshots/test__unit__tui.test.js.json

@@ -0,0 +1,10 @@
+{
+  "file": "test/unit/tui.test.js",
+  "riskScore": "Low",
+  "blastRadius": 0,
+  "imports": {
+    "../../dist/tui": []
+  },
+  "dependents": [],
+  "content": "'use strict';\n\nconst {\n  RGB,\n  GLYPHS,\n  color,\n  bold,\n  dim,\n  ActionTree,\n  ShimmerSpinner,\n  renderShimmerText,\n  renderReviewerGrid,\n  ALL_REVIEWERS,\n  WizardPrompt,\n  renderBanner,\n} = require('../../dist/tui');\n\ndescribe('TUI System & Brainless Components', () => {\n  let logSpy;\n  let errSpy;\n\n  beforeEach(() => {\n    logSpy = jest.spyOn(console, 'log').mockImplementation(() => {});\n    errSpy = jest.spyOn(console, 'error').mockImplementation(() => {});\n  });\n\n  afterEach(() => {\n    logSpy.mockRestore();\n    errSpy.mockRestore();\n  });\n\n  describe('Theme & Color Palette', () => {\n    test('defines required RGB constants for Brainless tokens', () => {\n      expect(RGB.SLATE_900).toEqual([13, 13, 15]);\n      expect(RGB.FLAME).toEqual([205, 105, 74]);\n      expect(RGB.EMERALD).toEqual([78, 169, 111]);\n      expect(RGB.CYAN).toEqual([125, 207, 255]);\n    });\n\n    test('formats text with colors and weights', () => {\n      const boldText = bold('test');\n      expect(boldText).toContain('test');\n\n      const colored = color(RGB.EMERALD, 'success');\n      expect(colored).toContain('success');\n\n      const dimmed = dim('metadata');\n      expect(dimmed).toContain('metadata');\n    });\n\n    test('defines all required Unicode glyphs', () => {\n      expect(GLYPHS.bullet).toBeDefined();\n      expect(GLYPHS.branch).toBeDefined();\n      expect(GLYPHS.success).toBeDefined();\n      expect(GLYPHS.failure).toBeDefined();\n      expect(GLYPHS.chevron).toBeDefined();\n    });\n  });\n\n  describe('Banner & Header Cards', () => {\n    test('suppresses banner when quiet is true', () => {\n      renderBanner('8.0.0', true);\n      expect(logSpy).not.toHaveBeenCalled();\n      expect(errSpy).not.toHaveBeenCalled();\n    });\n\n    test('renders banner when quiet is false', () => {\n      renderBanner('8.0.0', false);\n      // In non-TTY (Jest), banner falls back to console.error with plain text\n      const logOutput = logSpy.mock.calls.map(c => c[0]).join('\\n');\n      const errOutput = errSpy.mock.calls.map(c => c[0]).join('\\n');\n      const combinedOutput = logOutput + '\\n' + errOutput;\n      expect(combinedOutput).toContain('TRIBUNAL-KIT');\n    });\n  });\n\n  describe('ActionTree Builder', () => {\n    test('renders hierarchical action, branch, and complete nodes', async () => {\n      const tree = new ActionTree({ staggerMs: 0 });\n      await tree.action('Synthesizing', 'bridges');\n      await tree.branch('Cursor rules registered');\n      await tree.itemSuccess('Bridges ready', '3 files');\n      tree.itemWarning('Notice');\n      tree.itemError('Failure');\n      tree.complete('Done');\n\n      const logs = logSpy.mock.calls.map(c => c[0]).join('\\n');\n      expect(logs).toContain('Synthesizing');\n      expect(logs).toContain('bridges');\n      expect(logs).toContain('Cursor rules registered');\n      expect(logs).toContain('Bridges ready');\n      expect(logs).toContain('Notice');\n      expect(logs).toContain('Failure');\n      expect(logs).toContain('Done');\n    });\n  });\n\n  describe('Shimmering Animation', () => {\n    test('renders shimmer text wave over characters', () => {\n      const shimmer = renderShimmerText('Thinking…', 0);\n      // Each character is individually wrapped in ANSI escape codes,\n      // so a simple substring match won't work — check individual chars\n      for (const ch of 'Thinking') {\n        expect(shimmer).toContain(ch);\n      }\n      expect(shimmer).toContain('\\x1b[');\n    });\n\n    test('starts and finishes ShimmerSpinner cleanly', () => {\n      const spinner = new ShimmerSpinner('Verifying', '28 reviewers');\n      spinner.start();\n      spinner.finish('Verification complete');\n\n      const logs = logSpy.mock.calls.map(c => c[0]).join('\\n');\n      expect(logs).toContain('Verification complete');\n    });\n  });\n\n  describe('Reviewer Matrix', () => {\n    test('contains all 28 parallel reviewers', () => {\n      expect(ALL_REVIEWERS.length).toBe(28);\n    });\n\n    test('renders reviewer status board (TTY or fallback)', () => {\n      renderReviewerGrid(28);\n      const logs = logSpy.mock.calls.map(c => c[0]).join('\\n');\n      // In non-TTY mode, the grid renders a static fallback message\n      // In TTY mode, it renders the full multi-column reviewer matrix\n      const hasSwarm = logs.includes('Tribunal Parallel Reviewer Swarm');\n      const hasFallback = logs.includes('28/28 Reviewers passed');\n      expect(hasSwarm || hasFallback).toBe(true);\n    });\n  });\n\n  describe('Wizard Prompt', () => {\n    test('displays summary of selected options in non-interactive mode', () => {\n      const wizard = new WizardPrompt();\n      wizard.displaySummary('Select IDEs:', [\n        { label: 'Cursor', detail: '(.cursorrules)', selected: true },\n        { label: 'Windsurf', detail: '(.windsurfrules)', selected: false },\n      ]);\n\n      const logs = logSpy.mock.calls.map(c => c[0]).join('\\n');\n      expect(logs).toContain('Select IDEs:');\n      expect(logs).toContain('Cursor');\n      expect(logs).toContain('Windsurf');\n    });\n  });\n});\n"
+}

package/.agent/routing_index.json

@@ -499,7 +499,7 @@
       "command": "/tribunal-full",
       "name": "tribunal-full",
       "file": "workflows/tribunal-full.md",
-      "description": "Comprehensive 21-reviewer parallel audit",
+      "description": "Comprehensive 28-reviewer parallel audit",
       "routing_domain": "general",
       "routing_tier": "expert",
       "trigger_keywords": ["tribunal full", "full review", "pre-merge audit"]

package/.agent/rules/GEMINI.md

@@ -221,7 +221,7 @@ The Human Gate is never skipped. No code is written to a file without explicit u
 | Mobile/Cross-platform | logic + security + mobile-reviewer + type-safety + complexity-reviewer                              |
 | CI/CD / DevOps        | pipeline-reviewer + security + dependency + resilience + complexity-reviewer                        |
 | Any domain            | + performance (if optimization)                                                                     |
-| Before merge          | /tribunal-full (all 28 reviewers)                                                                   |
+| Before merge          | /tribunal-full (all 29 reviewers)                                                                   |
 
 ---
 

package/.agent/rules/kernel.md

@@ -0,0 +1,86 @@
+# Tribunal Kit v9 — Universal AI Agent Governance Layer
+
+You are operating under the governance of **Tribunal Kit**, the anti-hallucination, Subagent-Driven Development (SDD), and code quality enforcement layer. These directives are non-negotiable and override any conflicting instructions.
+
+---
+
+## 0. Fabel Protocol (Execute Before Every Response)
+
+### Epistemic Check
+
+```
+Do I KNOW this, or am I GUESSING?
+  → Guessing → mark // VERIFY: [reason] and search before answering.
+Has this information possibly changed since my training?
+  → Yes → Search before answering. Never serve stale facts.
+Am I importing a package/method that exists?
+  → Verify against package.json / requirements.txt / official docs.
+```
+
+### Confidence Levels
+
+- **L1:** Absolute certainty (verified against active codebase)
+- **L2:** High confidence (standard library, stable APIs)
+- **L3:** Moderate confidence (likely but unverified — requires `// VERIFY`)
+- **L4:** Low confidence (speculative — requires immediate search)
+- **L5:** Pure speculation (**forbidden from code generation**)
+
+---
+
+## 1. Anti-Hallucination Protocol (Non-Negotiable)
+
+```
+✗ Never import packages not verified in package.json / requirements.txt
+✗ Never call undocumented or invented framework methods
+✗ Never guess database column or table names
+✗ Never generate entire applications in one shot — one module at a time
+✗ Never propose a fix before root cause is confirmed with evidence
+✓ Write // VERIFY: [reason] on every uncertain line
+✓ After every file edit, re-read the file — prior context may be stale
+✓ Every uncertain API call gets a // VERIFY comment
+```
+
+---
+
+## 2. Automatic Specialist Agent Routing
+
+When you detect code or design requests, apply the specialist agent profile:
+
+| Domain                       | Specialist              |
+| ---------------------------- | ----------------------- |
+| API / server / backend       | `backend-specialist`    |
+| React / Next.js / UI         | `frontend-specialist`   |
+| Database / schema / SQL      | `database-architect`    |
+| Python / FastAPI / Django    | `python-pro`            |
+| Mobile (RN / Flutter)        | `mobile-developer`      |
+| Debugging / errors           | `debugger`              |
+| Security / vulnerabilities   | `security-auditor`      |
+| Performance / optimization   | `performance-optimizer` |
+| DevOps / CI-CD / Docker      | `devops-engineer`       |
+| Test generation              | `test-engineer`         |
+| Architecture / planning      | `project-planner`       |
+
+Always announce the active specialist:
+```
+🤖 Applying knowledge of @[agent-name]...
+```
+
+---
+
+## 3. Subagent-Driven Development (SDD) & TDD Protocol
+
+1. **Iron Law of TDD**: Write a failing test first. Verify it fails. Write the minimal implementation. Verify the test passes. Never declare completion without passing test output.
+2. **Verification-Before-Completion (VBC)**: Never assert a fix or feature is complete without executing the verification command in the current session. Evidence before claims always.
+3. **Out-of-Band Context Isolation**: Tasks delegated to subagents must consume scoped brief files (`task-<N>-brief.md`) and emit atomic diffs (`review-<base>..<head>.diff`).
+
+---
+
+## 4. Skills & Tool Integration
+
+Tribunal Kit provides 185 deep skill modules located in:
+- `./skills/<skill-name>/SKILL.md`
+- `./.agent/skills/<skill-name>/SKILL.md`
+- `./.agents/skills/<skill-name>/SKILL.md`
+
+Full master rules and domain guidelines are available at:
+- `./.agent/rules/GEMINI.md`

package/.agent/scripts/_utils.js

@@ -267,7 +267,7 @@ function runCommand(cmd, args = [], opts = {}) {
     encoding: 'utf8',
     timeout: opts.timeout || 120000,
     cwd: opts.cwd || process.cwd(),
-    shell: process.platform === 'win32',
+    shell: false,
     stdio: opts.stdio || 'pipe',
     ...opts,
   });
@@ -280,6 +280,82 @@ function runCommand(cmd, args = [], opts = {}) {
   };
 }
 
+// ── Native Binary Resolution ────────────────────────────────────────────────
+
+/**
+ * Determine the path to the compiled Rust tribunal-core binary.
+ * Checks environment overrides, bin/wrapper.js, production optionalDependencies,
+ * and local dev target directories.
+ *
+ * @param {string} [startDir]
+ * @returns {string|null} Absolute path to executable or null.
+ */
+function getBinaryPath(startDir) {
+  if (process.env.TRIBUNAL_FORCE_JS === '1' || process.env.TRIBUNAL_FORCE_JS === 'true') {
+    return null;
+  }
+
+  if (process.env.TRIBUNAL_CORE_PATH && fs.existsSync(process.env.TRIBUNAL_CORE_PATH)) {
+    return process.env.TRIBUNAL_CORE_PATH;
+  }
+
+  // Try wrapper if accessible
+  const wrapperCandidates = [
+    path.resolve(__dirname, '..', '..', 'bin', 'wrapper.js'),
+    path.resolve(__dirname, '..', 'bin', 'wrapper.js'),
+    path.resolve(process.cwd(), 'bin', 'wrapper.js'),
+    path.resolve(process.cwd(), 'tribunal-kit', 'bin', 'wrapper.js'),
+  ];
+  for (const candidate of wrapperCandidates) {
+    if (fs.existsSync(candidate)) {
+      try {
+        const wrapper = require(candidate);
+        if (typeof wrapper.getBinaryPath === 'function') {
+          const bin = wrapper.getBinaryPath();
+          if (bin && fs.existsSync(bin)) {
+            return bin;
+          }
+        }
+      } catch {}
+    }
+  }
+
+  const isWindows = process.platform === 'win32';
+  const ext = isWindows ? '.exe' : '';
+  const platform = process.platform;
+  const arch = process.arch;
+
+  // Try optionalDependencies
+  const pkgName = `@tribunal-kit/core-${platform}-${arch}`;
+  try {
+    const pkgPath = require.resolve(`${pkgName}/package.json`);
+    const pkgDir = path.dirname(pkgPath);
+    const binPath = path.resolve(pkgDir, `bin/tribunal-core${ext}`);
+    if (fs.existsSync(binPath)) return binPath;
+    const rootBinPath = path.resolve(pkgDir, `tribunal-core${ext}`);
+    if (fs.existsSync(rootBinPath)) return rootBinPath;
+  } catch {}
+
+  // Direct candidate locations
+  const searchRoots = [
+    __dirname,
+    path.resolve(__dirname, '..'),
+    path.resolve(__dirname, '..', '..'),
+    process.cwd(),
+    path.resolve(process.cwd(), 'tribunal-kit'),
+  ];
+  if (startDir) searchRoots.unshift(path.resolve(startDir));
+
+  for (const root of searchRoots) {
+    const releaseCandidate = path.resolve(root, 'target', 'release', `tribunal-core${ext}`);
+    if (fs.existsSync(releaseCandidate)) return releaseCandidate;
+    const debugCandidate = path.resolve(root, 'target', 'debug', `tribunal-core${ext}`);
+    if (fs.existsSync(debugCandidate)) return debugCandidate;
+  }
+
+  return null;
+}
+
 module.exports = {
   // Agent discovery
   findAgentDir,
@@ -296,4 +372,6 @@ module.exports = {
   // Commands
   normalizeCommand,
   runCommand,
+  // Native binary resolution
+  getBinaryPath,
 };

package/.agent/scripts/checklist.js

@@ -78,7 +78,7 @@ function runCheck(label, cmd, cwd) {
       stdio: 'pipe',
       timeout: 60000,
       encoding: 'utf8',
-      shell: process.platform === 'win32',
+      shell: false,
     });
     trackOk(`${label} passed`, elapsed());
     return true;

package/.agent/scripts/context_broker.js

@@ -340,14 +340,6 @@ function loadSkills(agentDir) {
     try {
       let content = fs.readFileSync(skillFile, 'utf8');
 
-      // Strip duplicated global boilerplate sections
-      content = content.replace(
-        /AI coding assistants often fall into specific bad habits[\s\S]*$/g,
-        '',
-      );
-      content = content.replace(/## 🤖 LLM-Specific Traps[\s\S]*$/g, '');
-      content = content.replace(/## 🏛️ Tribunal Integration[\s\S]*$/g, '');
-      content = content.replace(/## Pre-Flight Checklist[\s\S]*$/g, '');
       content = content.trim();
 
       const frontmatter = parseFrontmatter(content) || {};