tribunal-kit 7.0.0 → 9.0.0

package/.agent/skills/gsap-timeline/SKILL.md

@@ -2,8 +2,8 @@
 name: gsap-timeline
 description: Official GSAP skill for timelines — gsap.timeline(), position parameter, nesting, playback. Use when sequencing animations, choreographing keyframes, or when the user asks about animation sequencing, timelines, or animation order (in GSAP or when recommending a library that supports timelines).
 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:
   - motion-engineering
   - gsap-react
@@ -25,6 +25,12 @@ Before creating GSAP timelines, you MUST inspect:
 2. Timeline Defaults (Section 45) → Pass `{ defaults: { duration, ease } }` into `gsap.timeline()` constructor to avoid repetitive child declarations
 3. ScrollTrigger Placement (Section 101) → Place `scrollTrigger` ONLY on the top-level timeline, never on child tweens inside a timeline
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Official GSAP skill for timelines — gsap.timeline(), position parameter, nesting, playback. Use when sequencing animations, choreographing keyframes, or when the user asks about animation sequencing, timelines, or animation order (in GSAP or when recommending a library that supports timelines)..
+- **DO NOT activate when:** The task falls strictly outside gsap-timeline domain or belongs to a different dedicated specialist.
+
 ## When to Use This Skill
 
 Apply when building multi-step animations, coordinating several tweens in sequence or parallel, or when the user asks about timelines, sequencing, or keyframe-style animation in GSAP.
@@ -127,89 +133,17 @@ master.to('.c', { opacity: 0 }, '+=0.2');
 
 ---
 
-**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.
-
----
-
-## 🚨 LLM Trap Table
-
-| Pattern       | What AI Does Wrong            | What Is Actually Correct                                    |
-| :------------ | :---------------------------- | :---------------------------------------------------------- |
-| gsap-timeline | Position parameter as 2nd arg | Position parameter is the 3rd arg: tl.to(el, {x: 100}, "<") |
-| gsap-timeline | tl.add(gsap.to(...))          | tl.to(...) (just chain the tween directly)                  |
-| gsap-timeline | Hardcoded delays (delay: 1)   | Use position parameter (<, >, +=1) for relative timing      |
-
----
-
-## ✅ Pre-Flight Self-Audit
-
-Before producing any output, verify:
-
-```
-✅ Did I read the actual files before making claims about them?
-✅ Did I verify all method names against official GSAP documentation?
-✅ Did I add // VERIFY: on any uncertain API calls?
-✅ Are all imports from packages that actually exist in package.json?
-✅ Did I test my logic with edge cases (null, empty, 0, max)?
-✅ Did I avoid generating code for more than one module at a time?
-✅ Am I working from evidence, not assumption?
-```
-
----
-
-## 🔁 VBC Protocol (Verify → Build → Confirm)
-
-```
-VERIFY:  Read the actual codebase before writing anything
-BUILD:   Generate the smallest meaningful unit of code
-CONFIRM: Verify the output is correct before presenting
-```
-
-## 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?
@@ -218,8 +152,6 @@ 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/gsap-utils/SKILL.md

@@ -2,8 +2,8 @@
 name: gsap-utils
 description: Official GSAP skill for gsap.utils — clamp, mapRange, normalize, interpolate, random, snap, toArray, wrap, pipe. Use when the user asks about gsap.utils, clamp, mapRange, random, snap, toArray, wrap, or helper utilities in GSAP.
 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:
   - motion-engineering
   - gsap-react
@@ -25,6 +25,12 @@ Before using `gsap.utils`, you MUST inspect:
 2. Reusable Function Form (Section 22) → Omit the value argument (or pass `true` as 4th param for `random()`) to get a reusable function for high-frequency loops
 3. Component Scoping (Section 219) → Use `gsap.utils.selector(containerRef)` inside React components to restrict selector targets
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Official GSAP skill for gsap.utils — clamp, mapRange, normalize, interpolate, random, snap, toArray, wrap, pipe. Use when the user asks about gsap.utils, clamp, mapRange, random, snap, toArray, wrap, or helper utilities in GSAP..
+- **DO NOT activate when:** The task falls strictly outside gsap-utils domain or belongs to a different dedicated specialist.
+
 ## When to Use This Skill
 
 Apply when writing or reviewing code that uses **gsap.utils** for math, array/collection handling, unit parsing, or value mapping in animations (e.g. mapping scroll to a value, randomizing, snapping to a grid, or normalizing inputs).
@@ -304,89 +310,17 @@ https://gsap.com/docs/v3/HelperFunctions
 
 ---
 
-**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.
-
----
-
-## 🚨 LLM Trap Table
-
-| Pattern    | What AI Does Wrong                   | What Is Actually Correct                                           |
-| :--------- | :----------------------------------- | :----------------------------------------------------------------- |
-| gsap-utils | Using mapRange on units like "100px" | mapRange works only on pure numbers. Use getUnit/unitize           |
-| gsap-utils | Re-creating random per frame         | Pass true as 4th param: random(-10, 10, 1, true) for a reusable fn |
-| gsap-utils | Scoping global .box                  | Use gsap.utils.selector(container) inside components               |
-
----
-
-## ✅ Pre-Flight Self-Audit
-
-Before producing any output, verify:
-
-```
-✅ Did I read the actual files before making claims about them?
-✅ Did I verify all method names against official GSAP documentation?
-✅ Did I add // VERIFY: on any uncertain API calls?
-✅ Are all imports from packages that actually exist in package.json?
-✅ Did I test my logic with edge cases (null, empty, 0, max)?
-✅ Did I avoid generating code for more than one module at a time?
-✅ Am I working from evidence, not assumption?
-```
-
----
-
-## 🔁 VBC Protocol (Verify → Build → Confirm)
-
-```
-VERIFY:  Read the actual codebase before writing anything
-BUILD:   Generate the smallest meaningful unit of code
-CONFIRM: Verify the output is correct before presenting
-```
-
-## 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?
@@ -395,8 +329,6 @@ 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/harden/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: harden
 description: Make interfaces production-resilient with robust empty states, error boundaries, loading skeletons, offline indicators, and internationalization (i18n) layout support.
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - polish
   - baseline-ui
@@ -27,6 +27,12 @@ Before implementing UI component logic, you MUST inspect:
 
 Ensure UI components gracefully handle network failures, missing data, slow connections, extreme text lengths, and internationalization.
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Make interfaces production-resilient with robust empty states, error boundaries, loading skeletons, offline indicators, and internationalization (i18n) layout support..
+- **DO NOT activate when:** The task falls strictly outside harden domain or belongs to a different dedicated specialist.
+
 ---
 
 ## 5 Resilience Domains
@@ -62,28 +68,25 @@ Ensure UI components gracefully handle network failures, missing data, slow conn
 
 ---
 
-## 🤖 LLM-Specific Traps
+## 🏛️ Tribunal Verification & Guardrails
 
-1. **Swallowing Errors**: Catching API errors and displaying a blank container without informing the user.
-2. **Fixed Container Widths**: Using hardcoded `width: 200px` on containers with international text.
-
----
+**Slash command: `/review` or `/tribunal-full`**
+**Active reviewers: `logic-reviewer` · `security-auditor`**
 
-## 🏛️ Tribunal Integration (Anti-Hallucination)
-
-**Active reviewers: `resilience-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
-
 ```
-✅ Does every async data container have explicit loading, error, and empty states?
-✅ Are all containers built flexibly to accommodate 50% text expansion (i18n)?
-✅ Do error fallbacks include actionable retry buttons?
+✅ 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/harness-protocol/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: harness-protocol
 description: Rules and guidelines for the Marathon long-running agent harness
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - agentic-patterns
   - behavioral-modes
@@ -27,6 +27,12 @@ Before operating in Marathon long-running harness mode, you MUST inspect:
 
 This skill enforces the rules for the Marathon long-running agent harness.
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Rules and guidelines for the Marathon long-running agent harness.
+- **DO NOT activate when:** The task falls strictly outside harness-protocol domain or belongs to a different dedicated specialist.
+
 ## Rules
 
 1. Each session must start with `tk marathon init "spec"`
@@ -46,33 +52,17 @@ This skill enforces the rules for the Marathon long-running agent harness.
 
 ---
 
-## 🤖 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?
@@ -81,8 +71,6 @@ 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/i18n-localization/SKILL.md

@@ -2,8 +2,8 @@
 name: i18n-localization
 description: Internationalization (i18n) and localization mastery. Abstracting hardcoded strings, managing JSON/YAML translation dictionaries, bidirectional routing (RTL support for Arabic/Hebrew), Pluralization algorithms, date/currency formatting, and SSR locale detection in Next.js/React. Use when preparing an application for global multilingual scaling.
 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:
   - nextjs-react-expert
   - adapt
@@ -25,6 +25,12 @@ Before engineering multilingual or i18n features, you MUST inspect:
 2. Native `Intl` API Formatting (Section 71) → Use native `Intl.NumberFormat` and `Intl.DateTimeFormat` with active locale instead of third-party string parsers
 3. Logical CSS Properties for RTL (Section 94) → Use logical properties (`margin-inline-start`, `ms-4`) instead of physical `marginLeft`/`marginRight` for Arabic/Hebrew support
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Internationalization (i18n) and localization mastery. Abstracting hardcoded strings, managing JSON/YAML translation dictionaries, bidirectional routing (RTL support for Arabic/Hebrew), Pluralization algorithms, date/currency formatting, and SSR locale detection in Next.js/React. Use when preparing an application for global multilingual scaling..
+- **DO NOT activate when:** The task falls strictly outside i18n-localization domain or belongs to a different dedicated specialist.
+
 ## Hallucination Traps (Read First)
 
 - ❌ Concatenating translated strings (`'Hello ' + name`) -> ✅ Use interpolation: `t('greeting', { name })` to handle word order differences
@@ -34,8 +40,6 @@ Before engineering multilingual or i18n features, you MUST inspect:
 
 ---
 
-# i18n & Localization — Global Scale Mastery
-
 ---
 
 ## 1. The i18n Architecture (Next.js / React)
@@ -153,69 +157,17 @@ 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?
@@ -224,8 +176,6 @@ 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/impeccable/SKILL.md

@@ -1,8 +1,8 @@
 ---
 name: impeccable
 description: Flagship design engineering skill for creating production-grade, anti-generic frontend interfaces with supreme craftsmanship, visual hierarchy, typography, spatial systems, and micro-interactions.
-version: 3.0.0
-last-updated: 2026-07-30
+version: 4.0.0
+last-updated: 2026-09-07
 skills:
   - better-ui
   - better-colors
@@ -27,6 +27,12 @@ Before designing or engineering frontend components, you MUST inspect:
 
 The gold standard framework for crafting bespoke, world-class web applications that feel custom-built by elite design engineers.
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Flagship design engineering skill for creating production-grade, anti-generic frontend interfaces with supreme craftsmanship, visual hierarchy, typography, spatial systems, and micro-interactions..
+- **DO NOT activate when:** The task falls strictly outside impeccable domain or belongs to a different dedicated specialist.
+
 ---
 
 ## Core Pillars of Impeccable UI
@@ -59,30 +65,25 @@ The gold standard framework for crafting bespoke, world-class web applications t
 
 ---
 
-## 🤖 LLM-Specific Traps
+## 🏛️ Tribunal Verification & Guardrails
 
-1. **Generic AI Aesthetics**: Using purple-to-blue gradients, glow effects on every button, or heavy glassmorphism on white text.
-2. **Hardcoded Pixel Values**: Writing hardcoded `px` font sizes and dimensions instead of `rem` and fluid `clamp()` utilities.
-3. **Orphaned Text**: Failing to apply `text-wrap: balance` to multi-line headings.
-
----
+**Slash command: `/review` or `/tribunal-full`**
+**Active reviewers: `logic-reviewer` · `security-auditor`**
 
-## 🏛️ Tribunal Integration (Anti-Hallucination)
-
-**Active reviewers: `frontend-reviewer` · `ui-ux-auditor` · `motion-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
-
 ```
-✅ Are tabular numbers enabled on all changing numeric values?
-✅ Is heading text wrapped using `text-wrap: balance`?
-✅ Is container query responsiveness utilized where components are reused?
-✅ Are transitions tuned with custom cubic-bezier curves under 300ms?
+✅ 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/improve-codebase-architecture/SKILL.md

@@ -2,8 +2,8 @@
 name: improve-codebase-architecture
 description: Scans a codebase for deepening opportunities, architectural bottlenecks, tight coupling, and produces a prioritized visual improvement roadmap.
 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:
   - codebase-design
   - clean-code
@@ -25,10 +25,14 @@ Before refactoring or producing an architectural improvement roadmap, you MUST i
 2. Monolithic Controller Line Threshold (500 lines) (Section 27) → Flag any controller/service >500 lines with mixed concerns for immediate decomposition
 3. Incremental Refactoring Rule (Section 46) → Ban shotgun surgery across >5 files simultaneously; enforce phased, test-backed interface extractions
 
-# Improve Codebase Architecture — Architectural Audit & Refactoring
-
 Audit an existing codebase for structural rot, circular dependencies, monolithic controllers, and produce a prioritized refactoring roadmap.
 
+
+## Activation Boundaries
+
+- **Activate when:** Operating in tasks requiring Scans a codebase for deepening opportunities, architectural bottlenecks, tight coupling, and produces a prioritized visual improvement roadmap..
+- **DO NOT activate when:** The task falls strictly outside improve-codebase-architecture domain or belongs to a different dedicated specialist.
+
 ---
 
 ## 4 Audit Steps
@@ -57,27 +61,25 @@ Output audit findings in a clear prioritized Markdown table:
 
 ---
 
-## 🤖 LLM-Specific Traps
+## 🏛️ Tribunal Verification & Guardrails
 
-1. **Shotgun Surgery**: Refactoring 20 files at once without establishing tests or clean interfaces first.
-2. **Ignoring Existing Conventions**: Forgetting existing project patterns and forcing an incompatible framework structure.
+**Slash command: `/review` or `/tribunal-full`**
+**Active reviewers: `logic-reviewer` · `security-auditor`**
 
----
-
-## 🏛️ Tribunal Integration (Anti-Hallucination)
-
-**Active reviewers: `logic-reviewer` · `complexity-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
-
 ```
-✅ Are circular dependencies identified and resolved via clean seams?
-✅ Is the refactoring plan broken down into safe, testable phases?
+✅ 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.