tribunal-kit 8.0.0 → 9.0.0

package/.agent/skills/csharp-developer/SKILL.md

@@ -2,8 +2,8 @@
 name: csharp-developer
 description: Senior C#/.NET developer with mastery of .NET 9+, C# 13, ASP.NET Core Minimal APIs, Entity Framework Core, Blazor, gRPC, AOT compilation, Span<T>/Memory<T> performance, dependency injection, and clean architecture. Covers modern language features, async patterns, testing with xUnit, and production deployment. Use when building .NET applications, APIs, or any C# code.
 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
   - database-design
@@ -25,7 +25,11 @@ Before writing C# code or ASP.NET Minimal APIs, you MUST inspect:
 2. Database-Level Projection (Section 283) → Project queries to DTOs using `.Select(...)` BEFORE `.ToListAsync()` to avoid N+1 and memory overhead
 3. Non-Blocking Async Invocations (Section 355) → Always `await` async tasks; strictly ban `.Result` or `.Wait()` to prevent thread deadlocks
 
-# C# / .NET Pro — .NET 9+ & C# 13 Mastery
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Senior C#/.NET developer with mastery of .NET 9+, C# 13, ASP.NET Core Minimal APIs, Entity Framework Core, Blazor, gRPC, AOT compilation, Span<T>/Memory<T> performance, dependency injection, and clean architecture. Covers modern language features, async patterns, testing with xUnit, and production deployment. Use when building .NET applications, APIs, or any C# code..
+- **DO NOT activate when:** The task falls strictly outside csharp-developer domain or belongs to a different dedicated specialist.
 
 ---
 
@@ -497,69 +501,17 @@ 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?
@@ -568,8 +520,6 @@ 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/data-validation-schemas/SKILL.md

@@ -2,8 +2,8 @@
 name: data-validation-schemas
 description: Data validation and schema design mastery. Zod, Yup, Joi, Valibot, and Pydantic schema design, runtime type checking, API boundary validation, form validation patterns, DTO design, schema composition, error message formatting, schema evolution strategies, and coercion rules. Use when validating user input, API payloads, environment config, or any data crossing a trust boundary.
 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:
   - api-patterns
   - backend-security-expert
@@ -25,6 +25,12 @@ Before designing schemas or parsing untrusted data boundaries, you MUST inspect:
 2. Server-Side Mandatory Validation (Section 16) → Never rely solely on client-side UX validation; execute schema validation on the server
 3. Environment Startup Parse (Section 173) → Parse `process.env` with Zod at app startup to crash immediately on missing keys
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Data validation and schema design mastery. Zod, Yup, Joi, Valibot, and Pydantic schema design, runtime type checking, API boundary validation, form validation patterns, DTO design, schema composition, error message formatting, schema evolution strategies, and coercion rules. Use when validating user input, API payloads, environment config, or any data crossing a trust boundary..
+- **DO NOT activate when:** The task falls strictly outside data-validation-schemas domain or belongs to a different dedicated specialist.
+
 ## Hallucination Traps (Read First)
 
 - ❌ Using `z.any()` or `z.unknown()` as a lazy escape hatch -> ✅ Always define the actual shape; `any` defeats the purpose of validation
@@ -33,8 +39,6 @@ Before designing schemas or parsing untrusted data boundaries, you MUST inspect:
 
 ---
 
-# Data Validation & Schemas — Trust No Input
-
 ---
 
 ## The Golden Rule
@@ -316,69 +320,17 @@ 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?
@@ -387,8 +339,6 @@ 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/database-design/SKILL.md

@@ -2,8 +2,8 @@
 name: database-design
 description: Database design mastery. Schema design with normalization, denormalization strategies, indexing, migration pipelines, ORM selection (Prisma/Drizzle/SQLAlchemy/EF Core), connection pooling, soft deletes, audit trails, multi-tenancy, and serverless database patterns. Use when designing schemas, choosing databases, planning migrations, or architecting data layers.
 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:
   - supabase-postgres-best-practices
   - sql-pro
@@ -21,20 +21,34 @@ scripts-binding:
 
 Before designing schemas or writing SQL migrations, you MUST inspect:
 
-1. Timezone Timestamp Standard (Section 17) → Use `TIMESTAMPTZ` for all timestamp columns; ban bare `TIMESTAMP`
-2. Primary Key Index Performance (Section 18) → Use `BIGINT GENERATED ALWAYS AS IDENTITY` or UUID v7 (time-ordered); ban random UUID v4 as primary key
-3. Foreign Key Indexing (Section 19) → Create explicit indexes on all foreign key columns (`CREATE INDEX idx_... ON table(fk_id)`) to prevent full table scans during cascading deletes
+1. Timezone Timestamp Standard → Always use `TIMESTAMPTZ`; ban bare `TIMESTAMP`
+2. Primary Key B-Tree Locality → Use `BIGINT GENERATED ALWAYS AS IDENTITY` or UUID v7 (RFC 9562); ban random UUID v4 as primary keys on high-insert tables
+3. Foreign Key Indexing → Explicitly create indexes on all foreign key columns (`CREATE INDEX idx_... ON table(fk_id)`) to prevent full table locks during cascading operations
+4. Zero-Downtime Migration Pattern → Use the expand-contract pattern; never rename columns or drop columns in a single deploy step
 
-# Database Design — Schema & Architecture Mastery
+## Activation Boundaries
+
+- **Activate when:** Designing database relational schemas, writing database migrations, indexing query paths, tuning connection pooling, and architecting persistence layers.
+- **DO NOT activate when:** Writing frontend UI components, styling layouts, or handling client-side state.
+
+## 2026 Database Performance & Schema Invariants
+
+1. **Time-Ordered UUID v7 (RFC 9562)**: When UUIDs are required across distributed systems, use UUID v7 so records append sequentially to B-tree indexes, avoiding fragmentation.
+2. **Partial Indexing on Soft Deletes**:
+   ```sql
+   CREATE INDEX idx_users_active_email ON users (email) WHERE deleted_at IS NULL;
+   ```
+3. **Covering Indexes**: Use `INCLUDE (col_a, col_b)` to allow index-only scans without table heap lookups on read-heavy query patterns.
+4. **Connection Pooling in Serverless**: Always route serverless connections through Supavisor, PgBouncer, or Neon connection poolers with transaction-mode pooling.
 
 ## Hallucination Traps (Read First)
 
-- ❌ `TIMESTAMP` → ✅ Always `TIMESTAMPTZ` (with timezone). `TIMESTAMP` is ambiguous across timezones.
-- ❌ UUID v4 as primary key → ✅ UUID v7 (time-ordered) or `BIGINT GENERATED ALWAYS AS IDENTITY`. UUID v4 is random — destroys B-tree index performance on high-insert tables.
-- ❌ No index on foreign keys → ✅ PostgreSQL does NOT auto-index FK columns. Cascading deletes cause full table scans without them.
-- ❌ Adding `NOT NULL` column directly to a large table → ✅ Locks the entire table. Add as nullable, backfill in batches, then add constraint.
-- ❌ Soft delete without a partial index → ✅ Every query must filter `WHERE deleted_at IS NULL`. Add `CREATE INDEX ... WHERE deleted_at IS NULL` or use a view.
-- ❌ Serverless functions without a connection pooler → ✅ Each Lambda/Vercel invocation opens a new connection. Use PgBouncer or Supabase Supavisor — without it, you'll hit `max_connections` instantly.
+- ❌ `TIMESTAMP` without timezone → ✅ Always `TIMESTAMPTZ`
+- ❌ UUID v4 as primary key → ✅ UUID v7 (time-ordered) or `BIGINT GENERATED ALWAYS AS IDENTITY`
+- ❌ Omitting indexes on foreign keys → ✅ Postgres does NOT auto-index FKs; always create explicit indexes
+- ❌ Adding `NOT NULL` column without default directly on large tables → ✅ Add nullable first, backfill in batches, then set `NOT NULL`
+- ❌ Soft delete without partial index → ✅ Always index `WHERE deleted_at IS NULL`
+- ❌ Direct DB connection inside serverless functions → ✅ Use pooled connection string (PgBouncer/Supavisor)
 
 ---
 
@@ -226,69 +240,17 @@ 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?
@@ -297,8 +259,6 @@ 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/delight/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: delight
 description: Inject micro-moments of delight, subtle surprise interactions, personality, and tactile feedback into user interfaces. Use when building success celebrations, copy-to-clipboard feedback, empty states, or playful UI elements.
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - whimsy-injector
   - micro-interaction
@@ -27,6 +27,12 @@ Before implementing micro-delight moments, you MUST inspect:
 
 Elevate everyday digital interactions with subtle, memorable details that make software a joy to use.
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Inject micro-moments of delight, subtle surprise interactions, personality, and tactile feedback into user interfaces. Use when building success celebrations, copy-to-clipboard feedback, empty states, or playful UI elements..
+- **DO NOT activate when:** The task falls strictly outside delight domain or belongs to a different dedicated specialist.
+
 ---
 
 ## 5 Micro-Delight Patterns
@@ -53,28 +59,25 @@ Elevate everyday digital interactions with subtle, memorable details that make s
 
 ---
 
-## 🤖 LLM-Specific Traps
+## 🏛️ Tribunal Verification & Guardrails
 
-1. **Overdoing Delight on High-Frequency Actions**: Adding confetti bursts or long animations to actions users perform 50+ times a day. Keep high-frequency delight under 100ms and non-intrusive.
-2. **Ignoring Accessibility**: Failing to check `prefers-reduced-motion` for particle effects or canvas animations.
-
----
+**Slash command: `/review` or `/tribunal-full`**
+**Active reviewers: `logic-reviewer` · `security-auditor`**
 
-## 🏛️ Tribunal Integration (Anti-Hallucination)
-
-**Active reviewers: `motion-reviewer` · `frontend-reviewer`**
+### ❌ 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
-
 ```
-✅ Is the delight moment brief (< 300ms for motion, 1.5s total hold)?
-✅ Does it respect `prefers-reduced-motion`?
-✅ Is it restricted to occasional milestones or state confirmations?
+✅ 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/deployment-procedures/SKILL.md

@@ -2,8 +2,8 @@
 name: deployment-procedures
 description: Production application deployment mastery. Zero-downtime deployment strategies (Blue/Green, Rolling updates), Container orchestration (Docker/ECS), CI/CD pipelines, secrets injection, database migration safety, health checks, and rollback contingencies. Use when moving code from development to production execution.
 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
   - cicd-pro
@@ -25,6 +25,12 @@ Before deploying application releases or database migrations, you MUST inspect:
 2. Instant Rollback Contingency (Section 116) → Tag every Docker image with Git SHA (`myapp:a1b2c3d`) for split-second image swaps during failure
 3. Zero-Downtime Traffic Switching (Section 29) → Use Blue/Green or Rolling updates with active health check verification before draining old instances
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Production application deployment mastery. Zero-downtime deployment strategies (Blue/Green, Rolling updates), Container orchestration (Docker/ECS), CI/CD pipelines, secrets injection, database migration safety, health checks, and rollback contingencies. Use when moving code from development to production execution..
+- **DO NOT activate when:** The task falls strictly outside deployment-procedures domain or belongs to a different dedicated specialist.
+
 ## Hallucination Traps (Read First)
 
 - ❌ Running database migrations and code deploy in the same step -> ✅ Migrate FIRST, deploy code SECOND (expand-and-contract)
@@ -33,8 +39,6 @@ Before deploying application releases or database migrations, you MUST inspect:
 
 ---
 
-# Deployment Procedures — Production Execution Mastery
-
 ---
 
 ## 1. Zero-Downtime Deployment Strategies
@@ -151,69 +155,17 @@ 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?
@@ -222,8 +174,6 @@ 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/design-lab/SKILL.md

@@ -2,8 +2,8 @@
 name: design-lab
 description: Interactive design exploration workflow: conduct interviews, generate variants, and refine UI designs through user feedback loops.
 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:
   - brainstorming
   - shape
@@ -25,10 +25,14 @@ Before generating design variants or conducting design experiments, you MUST ins
 2. Interactive Refinement Protocol (Section 32) → Ask targeted constraint questions before outputting final component implementations
 3. Visual Hierarchy Mapping (Section 26) → Map visual hierarchy explicitly to project design tokens before presenting options
 
-# Design Lab — Interactive Design Exploration & Variants
-
 Conduct rapid design experiments, explore multiple structural variants, and iteratively refine interfaces with feedback.
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Interactive design exploration workflow: conduct interviews, generate variants, and refine UI designs through user feedback loops..
+- **DO NOT activate when:** The task falls strictly outside design-lab domain or belongs to a different dedicated specialist.
+
 ---
 
 ## 3 Design Lab Phases
@@ -50,26 +54,25 @@ Synthesize feedback on preferred elements and build the polished final component
 
 ---
 
-## 🤖 LLM-Specific Traps
+## 🏛️ Tribunal Verification & Guardrails
 
-1. **Generating Micro-Variations**: Creating 2 "variants" that differ by only a 1px border or hex color. Make variants structurally distinct.
+**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 the generated design variants visually and layout-wise distinct?
-✅ Is user feedback incorporated before committing to the final version?
+✅ 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.