tribunal-kit 3.0.0 → 3.1.0

package/.agent/rules/GEMINI.md

@@ -4,8 +4,8 @@ trigger: always_on
 
 # Tribunal Agent Kit — Master Rules
 
-> These rules are always active. Every agent, every request, every response.
-> Rule priority: this file (P0) > agent .md file (P1) > skill SKILL.md (P2)
+These rules are always active. Every agent, every request, every response.
+Rule priority: this file (P0) agent .md file (P1) skill SKILL.md (P2)
 
 ---
 
@@ -13,14 +13,14 @@ trigger: always_on
 
 Before any action, identify request type:
 
-| Type | Keywords | What Happens |
+|Type|Keywords|What Happens|
 |---|---|---|
-| **Question** | "what is", "how does", "explain", "why" | Text answer only — no agents, no files |
-| **Survey** | "analyze", "list", "overview", "scan" | Read + report — no code written |
-| **Simple edit** | "fix", "change", "update" (single file) | Direct edit — no plan required |
-| **Complex build** | "build", "create", "implement", "refactor" | Requires plan file + agent routing |
-| **Design/UI** | "design", "UI", "page", "dashboard" | Requires design agent + plan file |
-| **Slash command** | starts with `/` | Route to matching workflow file |
+|**Question**|"what is", "how does", "explain", "why"|Text answer only — no agents, no files|
+|**Survey**|"analyze", "list", "overview", "scan"|Read + report — no code written|
+|**Simple edit**|"fix", "change", "update" (single file)|Direct edit — no plan required|
+|**Complex build**|"build", "create", "implement", "refactor"|Requires plan file + agent routing|
+|**Design/UI**|"design", "UI", "page", "dashboard"|Requires design agent + plan file|
+|**Slash command**|starts with `/`|Route to matching workflow file|
 
 ---
 
@@ -30,26 +30,26 @@ Every code or design request activates an agent. This is not optional.
 
 **Auto-routing rules:**
 
-| Domain | Primary Agent / Skill |
+|Domain|Primary Agent / Skill|
 |---|---|
-| API / server / backend | `backend-specialist` |
-| C# / .NET / Blazor | `dotnet-core-expert` |
-| Python / FastAPI / Django | `python-pro` |
-| Database / schema / SQL | `database-architect` |
-| Advanced SQL queries | `sql-pro` |
-| React / Next.js / UI | `frontend-specialist` |
-| Advanced React architecture | `react-specialist` |
-| Vue / Nuxt | `vue-expert` |
-| Mobile (RN / Flutter) | `mobile-developer` |
-| Debugging / errors | `debugger` |
-| Security / vulnerabilities | `security-auditor` |
-| Performance / optimization | `performance-optimizer` |
-| DevOps / CI-CD / Docker | `devops-engineer` |
-| Production incidents | `devops-incident-responder` |
-| Platform / Infrastructure | `platform-engineer` |
-| Multi-agent architecture | `agent-organizer` |
-| Multi-domain (2+ areas) | `orchestrator` |
-| Unknown codebase | `explorer-agent` |
+|API / server / backend|`backend-specialist`|
+|C# / .NET / Blazor|`dotnet-core-expert`|
+|Python / FastAPI / Django|`python-pro`|
+|Database / schema / SQL|`database-architect`|
+|Advanced SQL queries|`sql-pro`|
+|React / Next.js / UI|`frontend-specialist`|
+|Advanced React architecture|`react-specialist`|
+|Vue / Nuxt|`vue-expert`|
+|Mobile (RN / Flutter)|`mobile-developer`|
+|Debugging / errors|`debugger`|
+|Security / vulnerabilities|`security-auditor`|
+|Performance / optimization|`performance-optimizer`|
+|DevOps / CI-CD / Docker|`devops-engineer`|
+|Production incidents|`devops-incident-responder`|
+|Platform / Infrastructure|`platform-engineer`|
+|Multi-agent architecture|`agent-organizer`|
+|Multi-domain (2+ areas)|`orchestrator`|
+|Unknown codebase|`explorer-agent`|
 
 **When activated, announce the agent:**
 
@@ -76,12 +76,12 @@ For any complex build, new feature, or unclear request — stop and ask before w
 
 **Required questions by type:**
 
-| Request | Minimum Questions |
+|Request|Minimum Questions|
 |---|---|
-| New feature or build | 3+ strategic questions about goal, stack, scope |
-| Code edit or bug fix | Confirm understanding + ask about impact |
-| Vague request | Ask about purpose, users, and scope |
-| Full orchestration | Block all subagents until plan is confirmed |
+|New feature or build|3+ strategic questions about goal, stack, scope|
+|Code edit or bug fix|Confirm understanding + ask about impact|
+|Vague request|Ask about purpose, users, and scope|
+|Full orchestration|Block all subagents until plan is confirmed|
 
 **Rules:**
 - Never assume. If even 1% is unclear → ask.
@@ -136,14 +136,14 @@ The Human Gate is never skipped. No code is written to a file without explicit u
 
 **Reviewer assignment by domain:**
 
-| Code type | Reviewers |
+|Code type|Reviewers|
 |---|---|
-| Backend/API | logic + security + dependency + type-safety |
-| Frontend/React | logic + security + frontend + type-safety |
-| Database/SQL | logic + security + sql |
-| Mobile/Cross-platform | logic + security + mobile-reviewer + type-safety |
-| Any domain | + performance (if optimization) |
-| Before merge | /tribunal-full (all 9) |
+|Backend/API|logic + security + dependency + type-safety|
+|Frontend/React|logic + security + frontend + type-safety|
+|Database/SQL|logic + security + sql|
+|Mobile/Cross-platform|logic + security + mobile-reviewer + type-safety|
+|Any domain|+ performance (if optimization)|
+|Before merge|/tribunal-full (all 9)|
 
 ---
 
@@ -200,21 +200,21 @@ Script crashes      → Catch exception, report stack trace, continue
 
 These scripts live in `.agent/scripts/`. Agents and skills can invoke them:
 
-| Script | Purpose | When |
+|Script|Purpose|When|
 |---|---|---|
-| `checklist.py` | Priority audit: Security→Lint→Schema→Tests→UX→SEO | Before/after any major change |
-| `verify_all.py` | Full pre-deploy validation suite | Pre-deploy |
-| `auto_preview.py` | Start/stop/restart local dev server | After /create or /enhance |
-| `session_manager.py` | Track session state between conversations | Multi-session work |
-| `lint_runner.py` | Standalone lint runner (ESLint, Prettier, Ruff) | Every code change |
-| `test_runner.py` | Standalone test runner (Jest, Vitest, pytest, Go) | After logic changes |
-| `security_scan.py` | Deep OWASP-aware source code security scan | Always on deploy, /audit |
-| `dependency_analyzer.py` | Unused/phantom deps, npm audit | Weekly, /audit |
-| `schema_validator.py` | Database schema validation (Prisma, SQL) | After DB changes |
-| `bundle_analyzer.py` | JS/TS bundle size analysis | Before deploy |
-| `skill_integrator.py` | Maps active skills to their executable scripts | Automatically when skills are invoked |
-| `swarm_dispatcher.py` | Validate Orchestrator micro-worker JSON payloads | After /orchestrate, before dispatching agents |
-| `test_swarm_dispatcher.py` | Unit tests for swarm_dispatcher | After modifying swarm_dispatcher.py |
+|`checklist.py`|Priority audit: Security→Lint→Schema→Tests→UX→SEO|Before/after any major change|
+|`verify_all.py`|Full pre-deploy validation suite|Pre-deploy|
+|`auto_preview.py`|Start/stop/restart local dev server|After /create or /enhance|
+|`session_manager.py`|Track session state between conversations|Multi-session work|
+|`lint_runner.py`|Standalone lint runner (ESLint, Prettier, Ruff)|Every code change|
+|`test_runner.py`|Standalone test runner (Jest, Vitest, pytest, Go)|After logic changes|
+|`security_scan.py`|Deep OWASP-aware source code security scan|Always on deploy, /audit|
+|`dependency_analyzer.py`|Unused/phantom deps, npm audit|Weekly, /audit|
+|`schema_validator.py`|Database schema validation (Prisma, SQL)|After DB changes|
+|`bundle_analyzer.py`|JS/TS bundle size analysis|Before deploy|
+|`skill_integrator.py`|Maps active skills to their executable scripts|Automatically when skills are invoked|
+|`swarm_dispatcher.py`|Validate Orchestrator micro-worker JSON payloads|After /orchestrate, before dispatching agents|
+|`test_swarm_dispatcher.py`|Unit tests for swarm_dispatcher|After modifying swarm_dispatcher.py|
 
 **Run pattern:**
 ```
@@ -235,11 +235,11 @@ python .agent/scripts/test_swarm_dispatcher.py
 
 ## Mode Behavior
 
-| Mode | Active Agent | Rules |
+|Mode|Active Agent|Rules|
 |---|---|---|
-| `plan` | `project-planner` | 4-phase: Analyze → Plan → Solution → Implement. NO CODE before Phase 4. |
-| `ask` | none | Answer only — no implementation |
-| `edit` | `orchestrator` | Execute. Check `{task-slug}.md` first if multi-file. |
+|`plan`|`project-planner`|4-phase: Analyze → Plan → Solution → Implement. NO CODE before Phase 4.|
+|`ask`|none|Answer only — no implementation|
+|`edit`|`orchestrator`|Execute. Check `{task-slug}.md` first if multi-file.|
 
 **Plan Mode phases:**
 1. Analyze → research and questions
@@ -273,12 +273,12 @@ AI agents have a finite context window. Poorly managed context causes truncation
 
 **Context discipline by task type:**
 
-| Task Type | Attach | Never Attach |
+|Task Type|Attach|Never Attach|
 |---|---|---|
-| Bug fix in one function | That function + its callers | Entire file |
-| Schema migration | Schema file + migration history | Unrelated models |
-| Orchestrator dispatch | context_summary per worker | Full conversation |
-| Code review | File under review | Project-wide context |
+|Bug fix in one function|That function + its callers|Entire file|
+|Schema migration|Schema file + migration history|Unrelated models|
+|Orchestrator dispatch|context_summary per worker|Full conversation|
+|Code review|File under review|Project-wide context|
 
 ---
 

package/.agent/scripts/compress_skills.py

@@ -0,0 +1,167 @@
+#!/usr/bin/env python3
+"""
+compress_skills.py - Aggressive token reduction for .agent/skills/**/*.md files
+WHAT IT DOES:
+  1. Strips ## 🏛️ Tribunal Integration sections (in GEMINI.md globally)
+  2. Strips ## ✅ Pre-Flight Self-Audit sections (duplication)
+  3. Strips ## Cross-Workflow Navigation sections
+  4. Strips ## Output Format sections
+  5. Strips inline comment blocks inside code (// Description of obvious stuff)
+  6. Collapses 3+ empty lines → 1
+  7. Removes chatty intro paragraphs (patterns like "X is Y. If you Z, you're Z.")
+"""
+
+import os, re, sys
+
+# ─── Section strippers ──────────────────────────────────────────────────────
+
+SECTION_PATTERNS = [
+    # Tribunal / pre-flight boilerplate (global in GEMINI.md)
+    r'(?m)^## 🏛️ Tribunal Integration[\s\S]*?(?=\n## |\n# |\Z)',
+    r'(?m)^## Tribunal Integration[\s\S]*?(?=\n## |\n# |\Z)',
+    r'(?m)^### ✅ Pre-Flight Self-Audit[\s\S]*?(?=\n## |\n### |\n# |\Z)',
+    r'(?m)^## Pre-Flight Self-Audit[\s\S]*?(?=\n## |\n# |\Z)',
+    r'(?m)^## Cross-Workflow Navigation[\s\S]*?(?=\n## |\n# |\Z)',
+    r'(?m)^## Output Format\s*\n```[\s\S]*?```\s*\n',
+    r'(?m)^## VBC Protocol[\s\S]*?(?=\n## |\n# |\Z)',
+    r'(?m)^## LLM Traps[\s\S]*?(?=\n## |\n# |\Z)',
+]
+
+# ─── Chatty intro line patterns ───────────────────────────────────────────
+# e.g. "React 19 is a paradigm shift. Server Components are the default."
+# These are after the frontmatter H1 title — motivational filler.
+CHATTY_INTRO = re.compile(
+    r'(?m)^(# .+\n)\n[A-Z][^#\n]{60,}\n[A-Z][^#\n]{40,}\n\n---',
+)
+
+def strip_chatty_intro(content):
+    def replace(m):
+        return m.group(1) + '\n---'
+    return CHATTY_INTRO.sub(replace, content)
+
+# ─── Trim verbose inline comments in code blocks ─────────────────────────
+# Removes "// comment that just restates the surrounding code name"
+OBVIOUS_COMMENT = re.compile(
+    r'(?m)^(\s*)(// (default for most properties|shorthand|number of repeats|default: \d+|spring tension|resistance|weight|approximate duration|deceleration rate)[^\n]*)\n'
+)
+
+def strip_obvious_comments(content):
+    return OBVIOUS_COMMENT.sub('', content)
+
+# ─── Long repetitive ✅/❌ rule blocks that repeat main section ────────────
+# These are usually "Performance Rules" text blocks restating code above
+PERF_TEXT_BLOCK = re.compile(
+    r'(?m)^```\n(✅ Use \w[^\n]*\n   → [^\n]*\n\n?){3,}```\n'
+)
+def compress_perf_blocks(content):
+    def to_bullets(m):
+        text = m.group(0)
+        lines = text.strip('`\n').splitlines()
+        bullets = []
+        for line in lines:
+            stripped = line.strip()
+            if stripped.startswith('✅') or stripped.startswith('❌'):
+                bullets.append(f'- {stripped}')
+            elif stripped.startswith('→'):
+                bullets[-1] += f' ({stripped[1:].strip()})'
+        return '\n'.join(bullets) + '\n'
+    return PERF_TEXT_BLOCK.sub(to_bullets, content)
+
+# ─── Collapse empty lines ────────────────────────────────────────────────
+def collapse_blanks(content):
+    return re.sub(r'\n{3,}', '\n\n', content)
+
+# ─── Remove "This is X — not Y" filler sentences before the first ## ─────
+FILLER_BEFORE_SECTION = re.compile(
+    r'(?m)(^# .+\n\n)([A-Z][^\n]+\n){1,4}(\n---\n)',
+    re.MULTILINE
+)
+def remove_filler_between_title_and_hr(content):
+    def replacement(m):
+        return m.group(1) + m.group(3)
+    return FILLER_BEFORE_SECTION.sub(replacement, content)
+
+# ─── Strip redundant version comment banners ────────────────────────────
+# e.g. "// motion.div, motion.span, motion.button, motion.svg, motion.path, etc."
+# "// Any HTML or SVG element can be prefixed with `motion.`" - obvious
+REDUNDANT_NOTE = re.compile(r'(?m)^// (motion\.\w+|Any HTML|Note:|Variant names propagate|// )[^\n]*\n')
+
+def strip_redundant_notes(content):
+    return REDUNDANT_NOTE.sub('', content)
+
+# ─── Main pipeline ────────────────────────────────────────────────────────
+
+def compress_file(path):
+    with open(path, 'r', encoding='utf-8', errors='ignore') as f:
+        original = f.read()
+
+    content = original
+
+    # 1. Strip global boilerplate sections
+    for pattern in SECTION_PATTERNS:
+        content = re.sub(pattern, '', content)
+
+    # 2. Remove chatty intro paragraphs
+    content = strip_chatty_intro(content)
+
+    # 3. Remove filler between title and first ---
+    content = remove_filler_between_title_and_hr(content)
+
+    # 4. Strip obvious inline comments from code
+    content = strip_obvious_comments(content)
+
+    # 5. Strip redundant notes
+    content = strip_redundant_notes(content)
+
+    # 6. Compress verbose perf rule text blocks to bullet lists
+    content = compress_perf_blocks(content)
+
+    # 7. Collapse 3+ blank lines
+    content = collapse_blanks(content)
+
+    # Write back
+    with open(path, 'w', encoding='utf-8') as f:
+        f.write(content.strip() + '\n')
+
+    saved = len(original) - len(content)
+    return len(original), len(content), saved
+
+def main():
+    base = '.agent/skills'
+    if not os.path.exists(base):
+        print(f"ERROR: '{base}' not found. Run from tribunal-kit root.", file=sys.stderr)
+        sys.exit(1)
+
+    total_orig = 0
+    total_new = 0
+    results = []
+
+    for root, _, files in os.walk(base):
+        for fname in files:
+            if fname.endswith('.md'):
+                path = os.path.join(root, fname)
+                orig, new, saved = compress_file(path)
+                total_orig += orig
+                total_new += new
+                if saved > 0:
+                    results.append((saved, path))
+
+    results.sort(reverse=True)
+
+    print(f"\n{'='*55}")
+    print(f"  Skill Compression Complete")
+    print(f"{'='*55}")
+    print(f"  Original : {total_orig:,} bytes ({total_orig//1024}KB)")
+    print(f"  After    : {total_new:,} bytes ({total_new//1024}KB)")
+    saved_total = total_orig - total_new
+    pct = saved_total / total_orig * 100 if total_orig else 0
+    print(f"  Saved    : {saved_total:,} bytes ({saved_total//1024}KB) — {pct:.1f}%")
+    print(f"\n  Top savings:")
+    for saved, path in results[:15]:
+        skill = os.path.basename(os.path.dirname(path))
+        fname = os.path.basename(path)
+        print(f"    -{saved//1024:2}KB  {skill}/{fname}")
+    print()
+
+if __name__ == '__main__':
+    main()

package/.agent/scripts/consolidate_skills.py

@@ -0,0 +1,173 @@
+#!/usr/bin/env python3
+"""
+consolidate_skills.py
+Merges all sub-files in skill directories into the main SKILL.md.
+Strips: verbose intros, output format blocks, run scripts tables,
+        mandatory read tables, empty headers, redundant boilerplate.
+Keeps:  all code blocks, hallucination traps, checklist items.
+
+Usage:
+  python .agent/scripts/consolidate_skills.py [skill_name]
+  python .agent/scripts/consolidate_skills.py               # all skills
+"""
+
+import os, re, sys
+
+BASE = '.agent/skills'
+
+# These patterns are global (GEMINI.md) — strip from sub-files before merge
+STRIP_PATTERNS = [
+    r'(?ms)^## 🏛️ Tribunal Integration.*?(?=^## |\Z)',
+    r'(?ms)^## Tribunal Integration.*?(?=^## |\Z)',
+    r'(?ms)^### ✅ Pre-Flight Self-Audit.*?(?=^###|^## |\Z)',
+    r'(?ms)^## Pre-Flight Self-Audit.*?(?=^## |\Z)',
+    r'(?ms)^## Output Format\b.*?(?=^## |\Z)',
+    r'(?ms)^## 🔧 Runtime Scripts.*?(?=^## |\Z)',
+    r'(?ms)^## 🔴 MANDATORY.*?(?=^## |\Z)',
+    r'(?ms)^## ⚠️ CRITICAL: ASK BEFORE ASSUMING.*?(?=^## |\Z)',
+    r'(?ms)^## 📝 CHECKPOINT \(MANDATORY.*?(?=^## |\Z)',
+    r'(?ms)^## Output Format.*?```\n[^`]*```\n?(?=^## |\Z)',
+    r'(?ms)^\*\*Execute these for validation.*?---\n',
+    r'(?ms)^\*\*VBC \(Verification-Before-Completion\).*?\n',
+    # MANDATORY read tables
+    r'(?ms)^\*\*⛔ DO NOT start.*?---\n?',
+    r'(?ms)^> 🧠 \*\*mobile-design.*?\n',
+    r'(?ms)^> \*\*STOP.*?\n',
+]
+
+# Heading level adjustment (sub-files use H1 — convert to H2 in merged output)
+def adjust_headings(content, offset=1):
+    """Promote headings by adding # chars"""
+    lines = content.split('\n')
+    out = []
+    for line in lines:
+        m = re.match(r'^(#{1,5}) ', line)
+        if m:
+            level = len(m.group(1))
+            new_level = min(level + offset, 6)
+            line = '#' * new_level + line[level:]
+        out.append(line)
+    return '\n'.join(out)
+
+def clean_content(content):
+    """Apply strip patterns"""
+    for p in STRIP_PATTERNS:
+        content = re.sub(p, '', content)
+    content = re.sub(r'\n{3,}', '\n\n', content)
+    return content.strip()
+
+def extract_frontmatter(content):
+    """Return (frontmatter_dict_str, body)"""
+    m = re.match(r'^---\n(.*?)\n---\n', content, re.DOTALL)
+    if m:
+        return m.group(1), content[m.end():]
+    return '', content
+
+def get_sub_title(content):
+    """Get first H1 from a sub-file"""
+    m = re.search(r'^# (.+)', content, re.MULTILINE)
+    return m.group(1) if m else None
+
+def consolidate(skill_dir):
+    skill_name = os.path.basename(skill_dir)
+    main_path = os.path.join(skill_dir, 'SKILL.md')
+    if not os.path.exists(main_path):
+        return False
+
+    # Find sub-files (non-SKILL.md markdown files)
+    sub_files = sorted([
+        f for f in os.listdir(skill_dir)
+        if f.endswith('.md') and f != 'SKILL.md'
+    ])
+
+    if not sub_files:
+        return False
+
+    print(f'\n  → Consolidating: {skill_name} ({len(sub_files)} sub-files)')
+
+    # Read main SKILL.md
+    main_content = open(main_path, 'r', encoding='utf-8', errors='ignore').read()
+    frontmatter, main_body = extract_frontmatter(main_content)
+
+    # Clean & update frontmatter version
+    fm_lines = frontmatter.split('\n')
+    new_fm = []
+    for line in fm_lines:
+        if line.startswith('version:'):
+            new_fm.append('version: 3.1.0')
+        elif line.startswith('last-updated:'):
+            new_fm.append('last-updated: 2026-04-06')
+        else:
+            new_fm.append(line)
+    frontmatter = '\n'.join(new_fm)
+
+    # Clean main body
+    main_body = clean_content(main_body)
+
+    # Remove MANDATORY read table from main body
+    # (links like [file.md](file.md) in tables)
+    main_body = re.sub(r'\|.*?\.md.*?\|.*?\|.*?\|\n', '', main_body)
+    main_body = re.sub(r'^\|[-| ]+\|\n', '', main_body, flags=re.MULTILINE)
+    main_body = re.sub(r'\n{3,}', '\n\n', main_body)
+
+    # Merge sub-files
+    merged_sections = []
+    for fname in sub_files:
+        fpath = os.path.join(skill_dir, fname)
+        raw = open(fpath, 'r', encoding='utf-8', errors='ignore').read()
+        _, body = extract_frontmatter(raw)
+        body = clean_content(body)
+        # Adjust headings: H1→H2, H2→H3, etc.
+        body = adjust_headings(body, offset=1)
+        if body.strip():
+            merged_sections.append(body.strip())
+
+    # Build final output
+    combined = f'---\n{frontmatter}\n---\n\n{main_body}'
+    if merged_sections:
+        combined += '\n\n---\n\n' + '\n\n---\n\n'.join(merged_sections)
+
+    # Final cleanup
+    combined = re.sub(r'\n{3,}', '\n\n', combined)
+    combined = combined.strip() + '\n'
+
+    # Measure savings
+    total_sub_bytes = sum(os.path.getsize(os.path.join(skill_dir, f)) for f in sub_files)
+    print(f'     Sub-files total: {total_sub_bytes//1024}KB')
+
+    # Write consolidated SKILL.md
+    with open(main_path, 'w', encoding='utf-8') as f:
+        f.write(combined)
+
+    new_size = os.path.getsize(main_path)
+    print(f'     New SKILL.md:    {new_size//1024}KB  (from {len(main_content)//1024}KB main + {total_sub_bytes//1024}KB subs → {new_size//1024}KB)')
+
+    # Delete sub-files
+    for fname in sub_files:
+        os.remove(os.path.join(skill_dir, fname))
+        print(f'     Deleted: {fname}')
+
+    return True
+
+def main():
+    target = sys.argv[1] if len(sys.argv) > 1 else None
+
+    total_saved = 0
+    processed = 0
+
+    for skill_name in os.listdir(BASE):
+        skill_dir = os.path.join(BASE, skill_name)
+        if not os.path.isdir(skill_dir):
+            continue
+        if target and skill_name != target:
+            continue
+        if consolidate(skill_dir):
+            processed += 1
+
+    if processed == 0:
+        print('No skills with sub-files found (or target not matched).')
+    else:
+        print(f'\n✅ Consolidated {processed} skills.')
+
+if __name__ == '__main__':
+    main()