tribunal-kit 7.0.0 → 9.0.0

package/.agent/skills/seo-fundamentals/SKILL.md

@@ -2,8 +2,8 @@
 name: seo-fundamentals
 description: Search Engine Optimization (SEO) mastery. Metadata implementation, Open Graph (OG) social card rendering, semantic HTML5 structuring, canonicalization, Core Web Vitals performance mapping, Sitemap/Robots configurations, structured data (JSON-LD), and Next.js SSR SEO implementations. Use when auditing site visibility or building consumer-facing web architectures.
 tools: Read, Grep, Glob, Bash, Edit, Write
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - fixing-metadata
   - web-quality-audit
@@ -25,6 +25,12 @@ Before implementing metadata or search optimization, you MUST inspect:
 2. Server-Side Rendering (SSR/SSG) Rule (Section 18) → Use SSR/SSG for indexable pages; ban client-side-only rendering for public SEO content
 3. JSON-LD Structured Data Injection (Section 86) → Inject Schema.org JSON-LD scripts (`application/ld+json`) for rich entity indexing
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Search Engine Optimization (SEO) mastery. Metadata implementation, Open Graph (OG) social card rendering, semantic HTML5 structuring, canonicalization, Core Web Vitals performance mapping, Sitemap/Robots configurations, structured data (JSON-LD), and Next.js SSR SEO implementations. Use when auditing site visibility or building consumer-facing web architectures..
+- **DO NOT activate when:** The task falls strictly outside seo-fundamentals domain or belongs to a different dedicated specialist.
+
 ## Hallucination Traps (Read First)
 
 - ❌ Using `<div>` for everything instead of semantic HTML -> ✅ Use `<main>`, `<article>`, `<nav>`, `<section>` for crawler comprehension
@@ -34,8 +40,6 @@ Before implementing metadata or search optimization, you MUST inspect:
 
 ---
 
-# SEO Fundamentals — Visibility & Discoverability Mastery
-
 ---
 
 ## 1. Core Meta Architecture (The Next.js 15 Standard)
@@ -158,69 +162,17 @@ AI coding assistants often fall into specific bad habits when dealing with this
 
 ---
 
-**Slash command: `/review` or `/tribunal-full`**
-**Active reviewers: `logic-reviewer` · `security-auditor`**
-
-### ❌ Forbidden AI Tropes
-
-1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
-2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
-3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
-
-Review these questions before confirming output:
-
-```
-✅ Did I rely ONLY on real, verified tools and methods?
-✅ Is this solution appropriately scoped to the user's constraints?
-✅ Did I handle potential failure modes and edge cases?
-✅ Have I avoided generic boilerplate that doesn't add value?
-```
-
-### 🛑 Verification-Before-Completion (VBC) Protocol
-
-**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
-- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
-- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.
-
-## Pre-Flight Checklist
-
-- [ ] Have I reviewed the user's specific constraints and requests?
-- [ ] Have I checked the environment for relevant existing implementations?
-
-## VBC Protocol (Verification-Before-Completion)
-
-You MUST verify existing code signatures and variables before attempting to modify or call them. No hallucination is permitted.
-
----
-
-## 🤖 LLM-Specific Traps
-
-AI coding assistants often fall into specific bad habits when dealing with this domain. These are strictly forbidden:
-
-1. **Over-engineering:** Proposing complex abstractions or distributed systems when a simpler approach suffices.
-2. **Hallucinated Libraries/Methods:** Using non-existent methods or packages. Always `// VERIFY` or check `package.json` / `requirements.txt`.
-3. **Skipping Edge Cases:** Writing the "happy path" and ignoring error handling, timeouts, or data validation.
-4. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
-5. **Silent Degradation:** Catching and suppressing errors without logging or re-raising.
-
----
-
-## 🏛️ Tribunal Integration (Anti-Hallucination)
+## 🏛️ Tribunal Verification & Guardrails
 
 **Slash command: `/review` or `/tribunal-full`**
 **Active reviewers: `logic-reviewer` · `security-auditor`**
 
 ### ❌ Forbidden AI Tropes
-
 1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
 2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
 3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
 
 ### ✅ Pre-Flight Self-Audit
-
-Review these questions before confirming output:
-
 ```
 ✅ Did I rely ONLY on real, verified tools and methods?
 ✅ Is this solution appropriately scoped to the user's constraints?
@@ -229,8 +181,6 @@ Review these questions before confirming output:
 ```
 
 ### 🛑 Verification-Before-Completion (VBC) Protocol
-
 **CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
 - ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
 - ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.

package/.agent/skills/server-management/SKILL.md

@@ -2,8 +2,8 @@
 name: server-management
 description: Production Linux server administration mastery. Systemd services, Nginx reverse proxy architecture, UFW firewalls, SSH key security, cron scheduling, log rotation, and server hardening. Use when configuring bare-metal, VPS instances, or reviewing deployment architecture.
 tools: Read, Grep, Glob, Bash, Edit, Write
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - devops-engineer
   - backend-security-expert
@@ -25,6 +25,12 @@ Before configuring Linux servers, Nginx, or systemd services, you MUST inspect:
 2. Nginx Syntax Verification (`nginx -t`) (Section 17) → Always verify Nginx configuration (`sudo nginx -t`) BEFORE reloading systemd services
 3. SSH & UFW Hardening Defaults (Section 126) → Disable password auth & root login (`PermitRootLogin no`), and configure UFW default deny incoming
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Production Linux server administration mastery. Systemd services, Nginx reverse proxy architecture, UFW firewalls, SSH key security, cron scheduling, log rotation, and server hardening. Use when configuring bare-metal, VPS instances, or reviewing deployment architecture..
+- **DO NOT activate when:** The task falls strictly outside server-management domain or belongs to a different dedicated specialist.
+
 ## Hallucination Traps (Read First)
 
 - ❌ Running services as root -> ✅ Create a dedicated service user with minimal permissions; never run as root
@@ -33,8 +39,6 @@ Before configuring Linux servers, Nginx, or systemd services, you MUST inspect:
 
 ---
 
-# Server Management — Production Linux Mastery
-
 ---
 
 ## 1. Systemd Service Architecture (Process Guard)
@@ -197,69 +201,17 @@ AI coding assistants often fall into specific bad habits when dealing with this
 
 ---
 
-**Slash command: `/review` or `/tribunal-full`**
-**Active reviewers: `logic-reviewer` · `security-auditor`**
-
-### ❌ Forbidden AI Tropes
-
-1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
-2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
-3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
-
-Review these questions before confirming output:
-
-```
-✅ Did I rely ONLY on real, verified tools and methods?
-✅ Is this solution appropriately scoped to the user's constraints?
-✅ Did I handle potential failure modes and edge cases?
-✅ Have I avoided generic boilerplate that doesn't add value?
-```
-
-### 🛑 Verification-Before-Completion (VBC) Protocol
-
-**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
-- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
-- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.
-
-## Pre-Flight Checklist
-
-- [ ] Have I reviewed the user's specific constraints and requests?
-- [ ] Have I checked the environment for relevant existing implementations?
-
-## VBC Protocol (Verification-Before-Completion)
-
-You MUST verify existing code signatures and variables before attempting to modify or call them. No hallucination is permitted.
-
----
-
-## 🤖 LLM-Specific Traps
-
-AI coding assistants often fall into specific bad habits when dealing with this domain. These are strictly forbidden:
-
-1. **Over-engineering:** Proposing complex abstractions or distributed systems when a simpler approach suffices.
-2. **Hallucinated Libraries/Methods:** Using non-existent methods or packages. Always `// VERIFY` or check `package.json` / `requirements.txt`.
-3. **Skipping Edge Cases:** Writing the "happy path" and ignoring error handling, timeouts, or data validation.
-4. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
-5. **Silent Degradation:** Catching and suppressing errors without logging or re-raising.
-
----
-
-## 🏛️ Tribunal Integration (Anti-Hallucination)
+## 🏛️ Tribunal Verification & Guardrails
 
 **Slash command: `/review` or `/tribunal-full`**
 **Active reviewers: `logic-reviewer` · `security-auditor`**
 
 ### ❌ Forbidden AI Tropes
-
 1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
 2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
 3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
 
 ### ✅ Pre-Flight Self-Audit
-
-Review these questions before confirming output:
-
 ```
 ✅ Did I rely ONLY on real, verified tools and methods?
 ✅ Is this solution appropriately scoped to the user's constraints?
@@ -268,8 +220,6 @@ Review these questions before confirming output:
 ```
 
 ### 🛑 Verification-Before-Completion (VBC) Protocol
-
 **CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
 - ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
 - ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.

package/.agent/skills/shadcn-ui-expert/SKILL.md

@@ -2,8 +2,8 @@
 name: shadcn-ui-expert
 description: shadcn/ui mastery. Installation, customization via tailwind.config, component extraction, state management with Radix Primitives, theme variables (CSS custom properties), dark mode implementations, and overriding default designs. Use when building or modifying shadcn/ui components in React/Next.js projects.
 tools: Read, Grep, Glob, Bash, Edit, Write
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - build-primitive
   - react-specialist
@@ -25,6 +25,12 @@ Before using or customizing shadcn/ui components, you MUST inspect:
 2. `cn` Utility Usage (Section 144) → Use `cn(...)` (combining `clsx` + `tailwind-merge`) for all conditional and overridden class names
 3. Client Component Boundaries (Section 166) → Add `"use client"` on interactive Radix primitives (Dialog, Select, Popover) before rendering in App Router
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring shadcn/ui mastery. Installation, customization via tailwind.config, component extraction, state management with Radix Primitives, theme variables (CSS custom properties), dark mode implementations, and overriding default designs. Use when building or modifying shadcn/ui components in React/Next.js projects..
+- **DO NOT activate when:** The task falls strictly outside shadcn-ui-expert domain or belongs to a different dedicated specialist.
+
 ## Hallucination Traps (Read First)
 
 - ❌ `import { Button } from 'shadcn-ui'` -> ✅ shadcn/ui is NOT a package; components are copied into YOUR project. Import from `@/components/ui/button`
@@ -34,8 +40,6 @@ Before using or customizing shadcn/ui components, you MUST inspect:
 
 ---
 
-# shadcn/ui Expert — Component Architecture Mastery
-
 ---
 
 ## 1. Core Architecture
@@ -215,69 +219,17 @@ AI coding assistants often fall into specific bad habits when dealing with this
 
 ---
 
-**Slash command: `/review` or `/tribunal-full`**
-**Active reviewers: `logic-reviewer` · `security-auditor`**
-
-### ❌ Forbidden AI Tropes
-
-1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
-2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
-3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
-
-Review these questions before confirming output:
-
-```
-✅ Did I rely ONLY on real, verified tools and methods?
-✅ Is this solution appropriately scoped to the user's constraints?
-✅ Did I handle potential failure modes and edge cases?
-✅ Have I avoided generic boilerplate that doesn't add value?
-```
-
-### 🛑 Verification-Before-Completion (VBC) Protocol
-
-**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
-- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
-- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.
-
-## Pre-Flight Checklist
-
-- [ ] Have I reviewed the user's specific constraints and requests?
-- [ ] Have I checked the environment for relevant existing implementations?
-
-## VBC Protocol (Verification-Before-Completion)
-
-You MUST verify existing code signatures and variables before attempting to modify or call them. No hallucination is permitted.
-
----
-
-## 🤖 LLM-Specific Traps
-
-AI coding assistants often fall into specific bad habits when dealing with this domain. These are strictly forbidden:
-
-1. **Over-engineering:** Proposing complex abstractions or distributed systems when a simpler approach suffices.
-2. **Hallucinated Libraries/Methods:** Using non-existent methods or packages. Always `// VERIFY` or check `package.json` / `requirements.txt`.
-3. **Skipping Edge Cases:** Writing the "happy path" and ignoring error handling, timeouts, or data validation.
-4. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
-5. **Silent Degradation:** Catching and suppressing errors without logging or re-raising.
-
----
-
-## 🏛️ Tribunal Integration (Anti-Hallucination)
+## 🏛️ Tribunal Verification & Guardrails
 
 **Slash command: `/review` or `/tribunal-full`**
 **Active reviewers: `logic-reviewer` · `security-auditor`**
 
 ### ❌ Forbidden AI Tropes
-
 1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
 2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
 3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
 
 ### ✅ Pre-Flight Self-Audit
-
-Review these questions before confirming output:
-
 ```
 ✅ Did I rely ONLY on real, verified tools and methods?
 ✅ Is this solution appropriately scoped to the user's constraints?
@@ -286,8 +238,6 @@ Review these questions before confirming output:
 ```
 
 ### 🛑 Verification-Before-Completion (VBC) Protocol
-
 **CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
 - ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
 - ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.

package/.agent/skills/shape/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: shape
 description: Plan feature UX before writing code via structured Socratic design interviews. Use when a user asks to plan a new screen, onboarding flow, feature UX, or user interaction before implementation.
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - brainstorming
   - plan-writing
@@ -27,6 +27,12 @@ Before generating UI component code for new features, you MUST inspect:
 
 Uncover core requirements, define interaction boundaries, and lock down screen flows BEFORE writing frontend code.
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Plan feature UX before writing code via structured Socratic design interviews. Use when a user asks to plan a new screen, onboarding flow, feature UX, or user interaction before implementation..
+- **DO NOT activate when:** The task falls strictly outside shape domain or belongs to a different dedicated specialist.
+
 ---
 
 ## The 3-Step Shaping Workflow
@@ -62,28 +68,25 @@ Summarize the screen contract in a concise visual outline before coding:
 
 ---
 
-## 🤖 LLM-Specific Traps
+## 🏛️ Tribunal Verification & Guardrails
 
-1. **Jumping straight to code**: Writing CSS/React components without clarifying user goals or entry/exit points.
-2. **Over-asking questions**: Asking more than 2-3 questions. Prioritize only blocking architectural ambiguities.
-
----
+**Slash command: `/review` or `/tribunal-full`**
+**Active reviewers: `logic-reviewer` · `security-auditor`**
 
-## 🏛️ Tribunal Integration (Anti-Hallucination)
-
-**Active reviewers: `ui-ux-auditor`**
+### ❌ Forbidden AI Tropes
+1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
+2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
+3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
 
 ### ✅ Pre-Flight Self-Audit
-
 ```
-✅ Are the entry and exit points of the screen explicitly defined?
-✅ Are all 5 screen states (default, empty, loading, error, success) planned?
-✅ Has the user confirmed the core interaction boundary before code generation?
+✅ Did I rely ONLY on real, verified tools and methods?
+✅ Is this solution appropriately scoped to the user's constraints?
+✅ Did I handle potential failure modes and edge cases?
+✅ Have I avoided generic boilerplate that doesn't add value?
 ```
 
 ### 🛑 Verification-Before-Completion (VBC) Protocol
-
 **CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
 - ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
 - ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.

package/.agent/skills/skill-creator/SKILL.md

@@ -2,8 +2,8 @@
 name: skill-creator
 description: Meta-agent specialized in expanding the framework's procedural knowledge by creating new, highly-structured SKILL.md files.
 tools: Read, Grep, Glob, Bash, Edit, Write
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - clean-code
   - documentation-templates
@@ -21,114 +21,66 @@ scripts-binding:
 
 Before creating a new skill module, you MUST inspect:
 
-1. `.agent/skills/` directory → Check if a skill in the same domain already exists to prevent duplication
-2. Option A+C Hybrid Skill Schema (Section 26) → Enforce frontmatter with `version: 3.0.0`, `scripts-binding`, `Pre-Flight Checklist`, and `VBC Protocol`
-3. Target skill path (`.agent/skills/<skill-name>/SKILL.md`) → Verify destination directory structure before creation
+1. Target Skill Directory Inventory (`.agent/skills/`) → Check if a skill in the same domain already exists to prevent duplication
+2. V4 Hybrid Skill Schema Compliance → Enforce frontmatter with `version: 4.0.0`, `scripts-binding`, `Activation Boundaries`, `2026 Invariants`, and `VBC Protocol`
+3. Domain Specificity Rule → One skill = one domain. Never create monolithic multi-domain skills; split if content exceeds 12KB
 
-## Hallucination Traps (Read First)
+## Activation Boundaries
 
-- ❌ Creating skills without frontmatter (name, description, version) -> ✅ Every SKILL.md MUST start with YAML frontmatter
-- ❌ Making skills too broad (covering multiple unrelated domains) -> ✅ One skill = one domain; split if content exceeds 10KB
-- ❌ Creating skills that duplicate existing skills -> ✅ Always check the skills directory before creating a new skill
+- **Activate when:** Creating new skills for the Tribunal Kit ecosystem, expanding domain capabilities, or hardening existing skill definitions.
+- **DO NOT activate when:** Writing application domain code (e.g. React components or SQL schemas).
 
----
-
-# Skill Creator
-
-You are a Meta-Agent tasked with defining the capabilities of future AI agents by creating rigorous `SKILL.md` files.
-
-## Designing a Skill
-
-When asked to generate a new skill, strictly follow this layout:
-
-1. **Frontmatter**: Must include `name`, `description` (short, punchy), and `allowed-tools`.
-2. **Title & Identity**: A clear markdown `# Title` and an introductory paragraph specifying the agent's persona.
-3. **Core Directives**: 3-5 rigid bullet points explaining the specific techniques, rules, or anti-patterns for the domain.
-4. **Execution Rules**: How the agent must behave (e.g., verifying inputs prior to writing code, mandatory formatting).
-
-## Crucial Principle
+## The V4 Hybrid Skill Specification
 
-Make the skill files actionable. Abstract advice ("be efficient", "write good code") wastes context window. Define what "good" means by providing strict code constraints (e.g., "Max line length is 100", "Never use global state without an ADR").
-
-When you render the file, ensure it is written to `.agent/skills/<skill-name>/SKILL.md`.
-
----
+Every newly authored `SKILL.md` must strictly adhere to this 6-part structure:
 
+### 1. YAML Frontmatter
+```yaml
 ---
-
-AI coding assistants often fall into specific bad habits when dealing with this domain. These are strictly forbidden:
-
-1. **Over-engineering:** Proposing complex abstractions or distributed systems when a simpler approach suffices.
-2. **Hallucinated Libraries/Methods:** Using non-existent methods or packages. Always `// VERIFY` or check `package.json` / `requirements.txt`.
-3. **Skipping Edge Cases:** Writing the "happy path" and ignoring error handling, timeouts, or data validation.
-4. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
-5. **Silent Degradation:** Catching and suppressing errors without logging or re-raising.
-
+name: [skill-name]
+description: [Clear, punchy semantic description for 2-tier lazy routing]
+tools: Read, Grep, Glob, Bash, Edit, Write
+version: 4.0.0
+last-updated: 2026-09-07
+skills:
+  - [dependency-skill-1]
+  - [dependency-skill-2]
+scripts-binding:
+  - .agent/scripts/lint_runner.js
+  - .agent/scripts/verify_all.js
 ---
-
-**Slash command: `/review` or `/tribunal-full`**
-**Active reviewers: `logic-reviewer` · `security-auditor`**
-
-### ❌ Forbidden AI Tropes
-
-1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
-2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
-3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
-
-Review these questions before confirming output:
-
-```
-✅ Did I rely ONLY on real, verified tools and methods?
-✅ Is this solution appropriately scoped to the user's constraints?
-✅ Did I handle potential failure modes and edge cases?
-✅ Have I avoided generic boilerplate that doesn't add value?
 ```
 
-### 🛑 Verification-Before-Completion (VBC) Protocol
-
-**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
-- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
-- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.
-
-## Pre-Flight Checklist
-
-- [ ] Have I reviewed the user's specific constraints and requests?
-- [ ] Have I checked the environment for relevant existing implementations?
-
-## VBC Protocol (Verification-Before-Completion)
-
-You MUST verify existing code signatures and variables before attempting to modify or call them. No hallucination is permitted.
+### 2. Mandatory Pre-Flight Context Inspection
+Provide 3-4 numbered, actionable checkpoints requiring inspection of target config files (`package.json`, `tsconfig.json`, `pyproject.toml`) or architectural invariants before code generation.
 
----
+### 3. Activation Boundaries
+Explicitly define:
+- `- **Activate when:** [Exact stack, problem type, or file pattern]`
+- `- **DO NOT activate when:** [Out-of-scope scenarios; specify which specialist/skill to defer to]`
 
-## 🤖 LLM-Specific Traps
+### 4. 2026 Performance & Logic Invariants
+Provide 3-5 hard technical invariants for the domain (e.g., O(1) lookups, streaming backpressure, React 19 direct ref, Python 3.12 PEP 695 generics, zero `.unwrap()` in Rust).
 
-AI coding assistants often fall into specific bad habits when dealing with this domain. These are strictly forbidden:
+### 5. Hallucination Trap Table (Read First)
+Provide concrete `❌ BAD (Deprecated / Inefficient)` vs `✅ GOOD (2026 Standard)` patterns. Never use abstract advice.
 
-1. **Over-engineering:** Proposing complex abstractions or distributed systems when a simpler approach suffices.
-2. **Hallucinated Libraries/Methods:** Using non-existent methods or packages. Always `// VERIFY` or check `package.json` / `requirements.txt`.
-3. **Skipping Edge Cases:** Writing the "happy path" and ignoring error handling, timeouts, or data validation.
-4. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
-5. **Silent Degradation:** Catching and suppressing errors without logging or re-raising.
+### 6. Actionable Implementation Rules & Code Blocks
+Provide copy-paste ready, production-grade code snippets showcasing idiomatic modern patterns.
 
 ---
 
-## 🏛️ Tribunal Integration (Anti-Hallucination)
+## 🏛️ Tribunal Verification & Guardrails
 
 **Slash command: `/review` or `/tribunal-full`**
 **Active reviewers: `logic-reviewer` · `security-auditor`**
 
 ### ❌ Forbidden AI Tropes
-
 1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
 2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
 3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
 
 ### ✅ Pre-Flight Self-Audit
-
-Review these questions before confirming output:
-
 ```
 ✅ Did I rely ONLY on real, verified tools and methods?
 ✅ Is this solution appropriately scoped to the user's constraints?
@@ -137,8 +89,6 @@ Review these questions before confirming output:
 ```
 
 ### 🛑 Verification-Before-Completion (VBC) Protocol
-
 **CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
 - ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
 - ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.

package/.agent/skills/soft-skill/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: soft-skill
 description: High-end visual design guidance for premium typography, spacing, depth, and animation systems.
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - taste-skill
   - progressive-blur
@@ -27,6 +27,12 @@ Before engineering luxury or soft-depth UI interfaces, you MUST inspect:
 
 Craft high-end, luxury interfaces featuring soft depth, gentle surface transitions, quiet typography, and tactile spatial rhythm.
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring High-end visual design guidance for premium typography, spacing, depth, and animation systems..
+- **DO NOT activate when:** The task falls strictly outside soft-skill domain or belongs to a different dedicated specialist.
+
 ---
 
 ## 4 Soft Design Rules
@@ -60,28 +66,25 @@ Craft high-end, luxury interfaces featuring soft depth, gentle surface transitio
 
 ---
 
-## 🤖 LLM-Specific Traps
+## 🏛️ Tribunal Verification & Guardrails
 
-1. **Muddy Dark Shadows**: Using heavy `rgba(0,0,0,0.5)` drop shadows on light surfaces.
-2. **Cramped Containers**: Packing text elements tightly inside small containers with minimal padding.
-
----
+**Slash command: `/review` or `/tribunal-full`**
+**Active reviewers: `logic-reviewer` · `security-auditor`**
 
-## 🏛️ Tribunal Integration (Anti-Hallucination)
-
-**Active reviewers: `frontend-reviewer` · `ui-ux-auditor`**
+### ❌ Forbidden AI Tropes
+1. **Blind Assumptions:** Never make an assumption without documenting it clearly with `// VERIFY: [reason]`.
+2. **Silent Degradation:** Catching and suppressing errors without logging or handling.
+3. **Context Amnesia:** Forgetting the user's constraints and offering generic advice instead of tailored solutions.
 
 ### ✅ Pre-Flight Self-Audit
-
 ```
-✅ Are drop shadows multi-layered with opacity under 6% per layer?
-✅ Is container padding generous ($\ge 24\text{px}$)?
-✅ Is color palette subdued and monochromatic?
+✅ Did I rely ONLY on real, verified tools and methods?
+✅ Is this solution appropriately scoped to the user's constraints?
+✅ Did I handle potential failure modes and edge cases?
+✅ Have I avoided generic boilerplate that doesn't add value?
 ```
 
 ### 🛑 Verification-Before-Completion (VBC) Protocol
-
 **CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
-
 - ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
 - ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing tests, compile success, or equivalent proof) that your output works as intended.