tribunal-kit 8.0.0 → 9.0.0

package/.agent/skills/opentelemetry-observability/SKILL.md

@@ -2,8 +2,8 @@
 name: opentelemetry-observability
 description: Full-stack distributed tracing, metrics, OpenTelemetry (OTel), Prometheus, Grafana Tempo, and zero-overhead observability instrumentation.
 tools: Read, Grep, Glob, Edit, Write
-version: 3.0.0
-last-updated: 2026-08-05
+version: 4.0.0
+last-updated: 2026-09-07
 script: .agent/scripts/test_runner.js
 scripts-binding:
   - .agent/scripts/test_runner.js
@@ -23,6 +23,12 @@ Before instrumenting applications:
 2. Trace Propagation → Propagate `traceparent` W3C headers across HTTP and message queues
 3. Sampling Policy → Implement head/tail sampling to reduce telemetry storage costs by 80%
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Full-stack distributed tracing, metrics, OpenTelemetry (OTel), Prometheus, Grafana Tempo, and zero-overhead observability instrumentation..
+- **DO NOT activate when:** The task falls strictly outside opentelemetry-observability domain or belongs to a different dedicated specialist.
+
 ## Custom Trace & Meter Instrumentation (TypeScript)
 
 ```typescript
@@ -53,7 +59,27 @@ export async function handleLogin(userId: string) {
 }
 ```
 
-## 🛑 Verification-Before-Completion (VBC) Protocol
+---
+
+## 🏛️ 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
+```
+✅ 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?
+```
 
-- Verify traces connect seamlessly from frontend click to DB query.
-- Confirm telemetry exporter overhead adds < 1ms latency to HTTP handlers.
+### 🛑 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/page-transition-animation/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: page-transition-animation
 description: Page and route transition patterns using the native View Transitions API, Framer Motion AnimatePresence, and Next.js App Router exit animations.
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - framer-motion-expert
   - 60fps-animation
@@ -27,6 +27,12 @@ Before engineering route or page transitions, you MUST inspect:
 
 Architect smooth, seamless page and route transitions without layout jumps or frozen exit states.
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Page and route transition patterns using the native View Transitions API, Framer Motion AnimatePresence, and Next.js App Router exit animations..
+- **DO NOT activate when:** The task falls strictly outside page-transition-animation domain or belongs to a different dedicated specialist.
+
 ---
 
 ## 3 Page Transition Architectures
@@ -93,28 +99,25 @@ export function RouteLayoutWrapper({ children }: { children: React.ReactNode })
 
 ---
 
-## 🤖 LLM-Specific Traps
+## 🏛️ Tribunal Verification & Guardrails
 
-1. **Broken Next.js App Router Exits**: Forgetting to set a unique `key={pathname}` on `AnimatePresence` children, causing exit animations to be skipped entirely.
-2. **Excessive Vertical Displacement**: Animating pages from `y: 100px` down, causing jarring vertical scroll shifts. Keep route shifts under `12px`.
-
----
+**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 `document.startViewTransition` feature-detected with fallback for older browsers?
-✅ Is `mode="wait"` set on `AnimatePresence` to prevent double-page DOM stacking?
-✅ Is vertical page movement subtle ($\le 12\text{px}$) during transitions?
+✅ 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/parallel-agents/SKILL.md

@@ -2,8 +2,8 @@
 name: parallel-agents
 description: Parallel processing coordination for multi-agent swarms. Asynchronous dispatches, merging divergent logic streams, race conditions in autonomous agents, avoiding Git conflicts in concurrent generation, and fan-out/fan-in processing patterns. Use when orchestrating multiple agents simultaneously.
 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:
   - agent-organizer
   - fabel-protocol
@@ -25,6 +25,12 @@ Before executing parallel agent dispatches, you MUST inspect:
 2. `Promise.allSettled()` error handling (Section 77) → Handle partial worker failures gracefully without crashing the pipeline
 3. Context Window Budget → Keep total context across all parallel workers under 80,000 tokens
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Parallel processing coordination for multi-agent swarms. Asynchronous dispatches, merging divergent logic streams, race conditions in autonomous agents, avoiding Git conflicts in concurrent generation, and fan-out/fan-in processing patterns. Use when orchestrating multiple agents simultaneously..
+- **DO NOT activate when:** The task falls strictly outside parallel-agents domain or belongs to a different dedicated specialist.
+
 ---
 
 ## Hallucination Traps (Read First)
@@ -35,8 +41,6 @@ Before executing parallel agent dispatches, you MUST inspect:
 
 ---
 
-# Parallel Agents — Concurrent Orchestration Mastery
-
 ---
 
 ## 1. Fan-Out / Fan-In Pattern
@@ -128,69 +132,17 @@ 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?
@@ -199,8 +151,6 @@ 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/performance-profiling/SKILL.md

@@ -2,8 +2,8 @@
 name: performance-profiling
 description: Performance profiling mastery. Core Web Vitals (LCP, CLS, INP), Lighthouse auditing, JavaScript profiling, React rendering optimization, bundle analysis, memory leak detection, database query profiling (EXPLAIN ANALYZE), load testing, and performance budgets. Use when optimizing performance, debugging slow pages, or establishing performance standards.
 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:
   - 60fps-animation
   - gsap-performance
@@ -25,7 +25,11 @@ Before optimizing performance or auditing benchmarks, you MUST inspect:
 2. Bundle Size Limits & Import Tree-Shaking (Section 55) → Enforce total gzipped JS $<200\text{KB}$; use deep imports (`lodash/debounce`) instead of full barrel imports
 3. SQL EXPLAIN ANALYZE Requirement (Section 156) → Profile database queries with `EXPLAIN ANALYZE` before making index or query optimization changes
 
-# Performance Profiling — Measurement-Driven Optimization
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Performance profiling mastery. Core Web Vitals (LCP, CLS, INP), Lighthouse auditing, JavaScript profiling, React rendering optimization, bundle analysis, memory leak detection, database query profiling (EXPLAIN ANALYZE), load testing, and performance budgets. Use when optimizing performance, debugging slow pages, or establishing performance standards..
+- **DO NOT activate when:** The task falls strictly outside performance-profiling domain or belongs to a different dedicated specialist.
 
 ---
 
@@ -248,69 +252,17 @@ 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?
@@ -319,8 +271,6 @@ 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/plan-writing/SKILL.md

@@ -2,8 +2,8 @@
 name: plan-writing
 description: Technical design and implementation planning mastery. Writing structured execution checklists, dependency mapping, establishing rollback protocols, segmenting monolithic tasks, writing ADRs (Architecture Decision Records), and defining verification criteria. Use when transitioning from ideation to coordinated 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:
   - project-planner
   - brainstorming
@@ -21,19 +21,34 @@ scripts-binding:
 
 Before writing implementation plans or execution blueprints, you MUST inspect:
 
-1. Verification Criteria Requirement (Section 15) → Include explicit "How to verify" criteria for every wave; ban plan steps without verification
-2. Phased Wave Segmentation (Section 41) → Segment multi-file tasks into testable, independent waves (Wave 1: Data, Wave 2: API, Wave 3: UI)
-3. Non-Destructive Defaults & Rollback Rules (Section 66) → Define fallback feature toggles and database rollback procedures prior to execution
+1. The Task Contract Requirement → Formulate Objective, Inputs, Outputs, Hard Constraints, and Acceptance Criteria before outlining file edits
+2. Verification Criteria Requirement → Include explicit "How to verify" commands (unit tests, curl, linter) for every execution step; ban unverified plan steps
+3. Phased Wave Segmentation → Segment multi-file tasks into testable, independent waves (Wave 1: Data, Wave 2: Logic/API, Wave 3: UI, Wave 4: Audit)
+4. Non-Destructive Defaults & Rollback Rules → Define fallback feature toggles and database migration rollback procedures prior to execution
 
-## Hallucination Traps (Read First)
+## Activation Boundaries
 
-- ❌ Writing plans without verification criteria -> ✅ Every plan needs a 'How to verify this worked' section
-- ❌ Planning at the wrong granularity (too high or too low) -> ✅ Plans should be at the component/feature level, not line-by-line or system-wide
-- ❌ Skipping the 'What could go wrong' section -> ✅ Identifying failure modes before implementation prevents costly rework
+- **Activate when:** Preparing complex multi-file architectural changes, writing implementation plans, establishing ADRs, segmenting large tasks into execution waves, and defining verification strategies.
+- **DO NOT activate when:** Executing small single-file typo fixes, simple CSS tweaks, or answering conceptual questions without code changes.
 
----
+## 2026 Plan Writing & Task Contract Invariants
 
-# Plan Writing — Execution Blueprints Mastery
+1. **The Strict Task Contract**:
+   Every plan must establish:
+   - **OBJECTIVE**: Precise outcome expected.
+   - **HARD CONSTRAINTS**: Unchangeable tech stack, performance budgets, backwards compatibility.
+   - **ACCEPTANCE CRITERIA**: Unambiguous, observable conditions proving success (e.g. `npm test exits 0`, `LCP < 1.2s`).
+2. **Chunking Limit (Max 5 Files Per Wave)**:
+   Never generate a single wave touching > 5 files. Break large epics into consecutive waves where each wave compiles and passes verification before proceeding.
+3. **Evidence-Based Closeout**:
+   Every wave ends with an automated command the agent or human must run to prove correctness before proceeding to the next wave.
+
+## Hallucination Traps (Read First)
+
+- ❌ Writing plans without verification criteria → ✅ Every plan needs a 'How to verify this worked' section
+- ❌ Planning at the wrong granularity (too high or too low) → ✅ Plans should be at the component/feature level
+- ❌ Skipping the 'What could go wrong' section → ✅ Identifying failure modes before implementation prevents costly rework
+- ❌ Multi-file mega plans without wave chunking → ✅ Break into independent, testable waves
 
 ---
 
@@ -45,9 +60,15 @@ Before altering multiple files or introducing a new system architecture, a rigid
 
 1. **Objective Context:** 2-sentence summary of the requested goal.
 2. **Architectural Handoff:** (What stack, what libraries, what constraints).
-3. **Dependency Tree Execution Order:** (Cannot build frontend UI until backend API exists).
-4. **File Blueprint:** Exact files expected to be touched (`[NEW] src/api/user.ts`, `[MODIFY] src/db/schema.prisma`).
-5. **Verification Protocol:** Exactly how the agent/human will prove the task is completed successfully.
+3. **Task-Level Interface Contracts (Mandatory for Subagent Decoupling):**
+   Every task MUST declare:
+   - `Consumes:` Exact function signatures and types it imports from prior tasks.
+   - `Produces:` Exact function signatures and types it exports for subsequent tasks.
+4. **The Zero-Placeholder Invariant:**
+   Never write "TBD", "TODO", "implement later", "add validation", or "write tests for above". Every step must contain complete, exact code blocks and verification commands.
+5. **Dependency Tree Execution Order:** (Cannot build frontend UI until backend API exists).
+6. **File Blueprint:** Exact files expected to be touched (`[NEW] src/api/user.ts`, `[MODIFY] src/db/schema.prisma`).
+7. **Verification Protocol:** Exactly how the agent/human will prove the task is completed successfully.
 
 ---
 
@@ -129,69 +150,17 @@ 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?
@@ -200,8 +169,6 @@ 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/platform-engineer/SKILL.md

@@ -2,8 +2,8 @@
 name: platform-engineer
 description: Platform Engineering and Internal Developer Portal (IDP) mastery. Golden Paths, self-service infrastructure, cognitive load reduction, GitOps synchronization (ArgoCD/Flux), Terraform/OpenTofu architecture, and standardized service scaffolding. Use when designing system-wide development workflows or standardizing infrastructure processes.
 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:
   - cloud-architect
   - devops-engineer
@@ -25,6 +25,12 @@ Before designing developer platforms or GitOps manifests, you MUST inspect:
 2. GitOps Continuous Reconciliation (Section 48) → Enforce declarative GitOps (`selfHeal: true`) via ArgoCD/Flux; ban manual `kubectl apply` or console modifications
 3. Higher-Level Terraform Modules (Section 76) → Encapsulate complex IaC logic into simple interface modules (`capacity`, `needs_replica`) to reduce cognitive load
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Platform Engineering and Internal Developer Portal (IDP) mastery. Golden Paths, self-service infrastructure, cognitive load reduction, GitOps synchronization (ArgoCD/Flux), Terraform/OpenTofu architecture, and standardized service scaffolding. Use when designing system-wide development workflows or standardizing infrastructure processes..
+- **DO NOT activate when:** The task falls strictly outside platform-engineer domain or belongs to a different dedicated specialist.
+
 ## Hallucination Traps (Read First)
 
 - ❌ Building internal platforms without talking to developers -> ✅ Platform engineering exists to reduce developer cognitive load; ask them what hurts
@@ -33,8 +39,6 @@ Before designing developer platforms or GitOps manifests, you MUST inspect:
 
 ---
 
-# Platform Engineering — Developer Experience Mastery
-
 ---
 
 ## 1. The "Golden Path" Architecture
@@ -127,69 +131,17 @@ 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?
@@ -198,8 +150,6 @@ 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/platform-engineering-opentofu/SKILL.md

@@ -2,8 +2,8 @@
 name: platform-engineering-opentofu
 description: Infrastructure as Code (IaC) with OpenTofu/Terraform/Pulumi, automated cloud provisioning, DevSecOps pipelines, and self-service platform engineering.
 tools: Read, Grep, Glob, Edit, Write
-version: 3.0.0
-last-updated: 2026-08-05
+version: 4.0.0
+last-updated: 2026-09-07
 script: .agent/scripts/verify_all.js
 scripts-binding:
   - .agent/scripts/security_scan.js
@@ -24,6 +24,12 @@ Before provisioning cloud infrastructure:
 2. Least Privilege IAM → Enforce strict role-based access control (RBAC) on all cloud resources
 3. Plan Validation → Run `tofu plan` and static security analysis (tfsec/checkov) before `tofu apply`
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Infrastructure as Code (IaC) with OpenTofu/Terraform/Pulumi, automated cloud provisioning, DevSecOps pipelines, and self-service platform engineering..
+- **DO NOT activate when:** The task falls strictly outside platform-engineering-opentofu domain or belongs to a different dedicated specialist.
+
 ## Production AWS VPC & ECS Module Blueprint
 
 ```hcl
@@ -62,7 +68,27 @@ resource "aws_subnet" "public_a" {
 }
 ```
 
-## 🛑 Verification-Before-Completion (VBC) Protocol
+---
+
+## 🏛️ 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
+```
+✅ 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?
+```
 
-- Run `tofu validate` and security scan before applying changes.
-- Ensure rollback plan is explicitly documented.
+### 🛑 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.