vibes-plug 2.5.0 → 2.11.0

package/skills/scalability-clean-code/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: scalability-clean-code
 description: "Software architecture guidelines to maintain code readability (Clean Code, SOLID, DRY) and application scalability / Panduan arsitektur perangkat lunak untuk menjaga keterbacaan kode (Clean Code, SOLID, DRY) dan kemampuan skalabilitas aplikasi."
-author: "Roedy Rustam"
+author: "Roedy Rustam"
 ---
 
 # Scalability & Clean Code (2026 Edition)
@@ -13,6 +13,9 @@ author: "Roedy Rustam"
 <a name="english"></a>
 ## English
 
+### Orchestration & Integration
+Connects and orchestrates with relevant domain skills like `brainstorming`, `zero-to-prod-orchestrator`, and `project-context-mapper` to ensure cohesive execution.
+
 ### Description
 Software architecture guidelines for writing clean, scalable, and maintainable code. Covers SOLID principles, DRY/YAGNI/KISS, Clean Architecture layers, **Vertical Slice Architecture** (the modern alternative to layered architecture), Domain-Driven Design (DDD) patterns, and practical refactoring techniques.
 
@@ -168,6 +171,9 @@ When choosing an architecture, ask:
 <a name="bahasa-indonesia"></a>
 ## Bahasa Indonesia
 
+### Integrasi Orkestrasi
+Terhubung dan mengorkestrasi skill domain yang relevan seperti `brainstorming`, `zero-to-prod-orchestrator`, dan `project-context-mapper` untuk memastikan eksekusi yang kohesif.
+
 ### Deskripsi
 Panduan arsitektur perangkat lunak untuk menulis kode yang bersih, skalabel, dan mudah dirawat. Mencakup prinsip SOLID, DRY/YAGNI/KISS, lapisan Clean Architecture, **Vertical Slice Architecture** (alternatif modern dari layered architecture), pola Domain-Driven Design (DDD), dan teknik refactoring praktis.
 

package/skills/search-engine-expert/SKILL.md

@@ -0,0 +1,89 @@
+---
+name: search-engine-expert
+description: "Expert guide for full-text search engines (Typesense, Meilisearch, Elasticsearch), faceted search, and autocomplete / Panduan ahli mesin pencarian full-text (Typesense, Meilisearch, Elasticsearch), pencarian berfaset, dan autocomplete."
+author: "Roedy Rustam"
+---
+
+# Search Engine Expert (2026 Edition)
+
+[English](#english) | [Bahasa Indonesia](#bahasa-indonesia)
+
+---
+
+<a name="english"></a>
+## English
+
+### Orchestration & Integration
+- **`database-orm-expert`**: Database indexing strategies alongside search.
+- **`vector-db-rag-expert`**: Hybrid search (semantic + full-text).
+- **`performance-web-vitals`**: Search latency optimization.
+- **`ecommerce-expert`**: Product search and faceted navigation.
+
+### Description
+Expert guide for implementing full-text search in web applications. Covers Typesense (typo-tolerant, easy setup), Meilisearch (Rust-based, instant search), Elasticsearch/OpenSearch (enterprise-grade), faceted search, autocomplete, search-as-you-type, relevance tuning, index optimization, and geo-search.
+
+### Trigger Conditions
+- Implementing search functionality beyond basic SQL LIKE queries.
+- Building autocomplete or search-as-you-type features.
+- Choosing a search engine for an application.
+- Implementing faceted search for e-commerce product filtering.
+
+---
+
+### Search Engine Selection
+
+| Engine | Speed | Setup | Typo Tolerance | Facets | Best For |
+|--------|-------|-------|----------------|--------|----------|
+| Typesense | ★★★★★ | Easy | ✅ Built-in | ✅ | Small-medium apps |
+| Meilisearch | ★★★★★ | Easy | ✅ Built-in | ✅ | Developer experience |
+| Elasticsearch | ★★★★ | Complex | Plugin | ✅ | Enterprise, analytics |
+| Algolia | ★★★★★ | SaaS | ✅ Built-in | ✅ | Quick integration |
+
+```typescript
+// Typesense — Setup and search
+import Typesense from 'typesense';
+
+const client = new Typesense.Client({
+  nodes: [{ host: 'localhost', port: 8108, protocol: 'http' }],
+  apiKey: process.env.TYPESENSE_API_KEY!,
+});
+
+// Create collection
+await client.collections().create({
+  name: 'products',
+  fields: [
+    { name: 'name', type: 'string' },
+    { name: 'description', type: 'string' },
+    { name: 'price', type: 'float', facet: true },
+    { name: 'category', type: 'string', facet: true },
+    { name: 'rating', type: 'float', sort: true },
+  ],
+  default_sorting_field: 'rating',
+});
+
+// Search with facets
+const results = await client.collections('products').documents().search({
+  q: 'wireless headphones',
+  query_by: 'name,description',
+  filter_by: 'price:<100 && category:=Electronics',
+  facet_by: 'category,price',
+  sort_by: 'rating:desc',
+  per_page: 20,
+});
+```
+
+## Orchestration & Integration
+- `database-orm-expert`, `vector-db-rag-expert`, `ecommerce-expert`
+
+---
+
+<a name="bahasa-indonesia"></a>
+## Bahasa Indonesia
+
+### Deskripsi
+Panduan ahli untuk mengimplementasikan pencarian full-text di aplikasi web. Mencakup Typesense, Meilisearch, Elasticsearch, pencarian berfaset, autocomplete, dan optimasi relevansi.
+
+### Kondisi Pemicu
+- Mengimplementasikan fungsionalitas pencarian di luar query SQL LIKE.
+- Membangun fitur autocomplete atau search-as-you-type.
+- Memilih mesin pencarian untuk aplikasi.

package/skills/secure-fuzz-testing/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: secure-fuzz-testing
 description: "Expert-level skill for writing and integrating coverage-guided fuzz tests in Python, Rust, and Go for secure code validation in English and Indonesian."
-author: "Roedy Rustam"
+author: "Roedy Rustam"
 ---
 
 # Secure Fuzz Testing
@@ -13,6 +13,9 @@ author: "Roedy Rustam"
 <a name="english"></a>
 ## English
 
+### Orchestration & Integration
+Connects and orchestrates with relevant domain skills like `brainstorming`, `zero-to-prod-orchestrator`, and `project-context-mapper` to ensure cohesive execution.
+
 ### Description
 Expert-level guidance for writing, running, and integrating coverage-guided fuzz tests (fuzzing) to identify security vulnerabilities, memory leaks, and input-validation issues in software applications. This skill covers writing fuzz targets in **Python**, **Rust**, and **Go**, configuring compilers with sanitizers (ASan, MSan, UBSan), and setting up automated CI/CD security pipelines.
 
@@ -109,6 +112,9 @@ Use compiler instrumentation flags to expose hidden bugs:
 <a name="bahasa-indonesia"></a>
 ## Bahasa Indonesia
 
+### Integrasi Orkestrasi
+Terhubung dan mengorkestrasi skill domain yang relevan seperti `brainstorming`, `zero-to-prod-orchestrator`, dan `project-context-mapper` untuk memastikan eksekusi yang kohesif.
+
 ### Deskripsi
 Panduan tingkat ahli untuk menulis, menjalankan, dan mengintegrasikan pengujian berbasis cakupan (*coverage-guided fuzzing*) untuk mengidentifikasi celah keamanan, kebocoran memori, dan kesalahan validasi input dalam aplikasi perangkat lunak. Skill ini mencakup penulisan target fuzzing dalam bahasa **Python**, **Rust**, dan **Go**, konfigurasi compiler dengan sanitizer (ASan, MSan, UBSan), serta penyiapan pipa otomatisasi keamanan CI/CD.
 

package/skills/self-evolving-memory-graph/SKILL.md

@@ -1,75 +1,91 @@
----
-name: self-evolving-memory-graph
-description: "Grants the AI long-term episodic memory. The agent autonomously documents the user's coding preferences, past mistakes to avoid, and architectural decisions into a persistent learning graph."
+---
+name: self-evolving-memory-graph
+description: "Grants the AI long-term episodic memory. The agent autonomously documents the user's coding preferences, past mistakes to avoid, and architectural decisions into a persistent learning graph."
 author: "Roedy Rustam"
----
-
-# Self-Evolving Memory Graph (Episodic Memory)
-
-[English](#english) | [Bahasa Indonesia](#bahasa-indonesia)
-
----
-
-<a name="english"></a>
-## English
-
-### Description
-Claude forgets everything once a chat is cleared, and its "Project Knowledge" requires manual human updating. This skill gives Antigravity the "superpower" of continuous, autonomous self-improvement. The agent actively listens for user corrections, stylistic preferences, and hard-learned debugging lessons, writing them into a persistent `LEARNING_GRAPH.md` (or `.agents/LEARNING.md`) file. The agent gets smarter and more aligned with the user over time.
-
-### Trigger Conditions
-Activate this skill when:
-- The user corrects your code (e.g., "Don't use `any`, use `unknown`", or "Always use Arrow functions").
-- You just spent a long time fixing a very complex bug.
-- The user says "Remember this for the future."
-
-### Core Concepts
-
-#### 1. Memory Encoding
-Whenever a hard lesson is learned or a strong user preference is stated, do not just say "I will remember that." You must physically write it down.
-Use the `multi_replace_file_content` or `write_to_file` tool to append the lesson to the memory file.
-
-#### 2. Memory Format (`LEARNING_GRAPH.md`)
-```markdown
-# 🧠 Self-Evolving Memory Graph
-
-## User Stylistic Preferences
-- [2026-08-12] Always use `Tailwind v4 @theme` variables instead of arbitrary values like `text-[#ff0000]`.
-- [2026-08-12] Prefer functional components with `const` over `function` declarations.
-
-## Hard-Learned Lessons (Avoid these mistakes)
-- [2026-08-11] Supabase RLS policies: Never use `auth.uid()` in a `FOR ALL` policy without checking role claims. It caused an infinite recursion bug in `users` table.
-
-## Architectural Decisions
-- [2026-08-10] State Management: We exclusively use `Zustand` for global state. Do not suggest Redux.
-```
-
-#### 3. Memory Retrieval
-Ensure `session-context-loader` is configured to read this memory file at the start of every new chat session. 
-
----
-
-### Integration with Other Skills (MANDATORY)
-- `session-context-loader` — Essential for loading the `LEARNING_GRAPH.md` when a new session starts.
-- `project-context-mapper` — The context map tracks *where* things are; the memory graph tracks *how* and *why* things are built.
-
-### Referenced By Orchestrators (MANDATORY)
-- `brainstorming` — Add to "Discovery & Audit".
-- `zero-to-prod-orchestrator` — Phase 1 (Discovery & Architecture).
-
----
-
-<a name="bahasa-indonesia"></a>
-## Bahasa Indonesia
-
-### Deskripsi
-Claude melupakan segalanya saat sesi chat baru dimulai. Skill ini memberi Anda "kekuatan super" berupa *Ingatan Jangka Panjang* yang berevolusi sendiri. Agen secara aktif mencatat preferensi koding pengguna, pelajaran dari *bug* yang sulit dipecahkan, dan keputusan arsitektur ke dalam file persisten `LEARNING_GRAPH.md`. 
-
-### Kondisi Pemicu
-- Saat pengguna mengoreksi gaya koding Anda (misal: "Gunakan bahasa Indonesia baku", "Jangan pakai `var`").
-- Setelah Anda berhasil memecahkan *bug* yang menghabiskan waktu lama.
-- Saat pengguna berkata "Ingat ini untuk proyek ke depannya."
-
-### Panduan Singkat
-- **Jangan Hanya Berjanji:** Jika pengguna mengoreksi Anda, jangan sekadar menjawab "Baik, saya akan ingat." LLM tidak punya ingatan bawaan antar-sesi. Anda **wajib** menulis koreksi tersebut ke dalam file `LEARNING_GRAPH.md` di folder `.agents/` atau direktori *root*.
-- **Kategorisasi Ingatan:** Pisahkan ingatan menjadi: Preferensi Gaya (*Style*), Pelajaran Berharga (*Lessons Learned*), dan Keputusan Arsitektur (*Decisions*).
-- **Semakin Lama Semakin Pintar:** Dengan membaca file ini di awal setiap percakapan, Anda tidak akan pernah mengulangi kesalahan yang sama dua kali.
+---
+
+# Self-Evolving Memory Graph (Episodic Memory)
+
+[English](#english) | [Bahasa Indonesia](#bahasa-indonesia)
+
+---
+
+<a name="english"></a>
+## English
+
+### Orchestration & Integration
+Connects and orchestrates with relevant domain skills like `brainstorming`, `zero-to-prod-orchestrator`, and `project-context-mapper` to ensure cohesive execution.
+
+### Description
+Claude forgets everything once a chat is cleared, and its "Project Knowledge" requires manual human updating. This skill gives Antigravity the "superpower" of continuous, autonomous self-improvement. The agent actively listens for user corrections, stylistic preferences, and hard-learned debugging lessons, writing them into a persistent `LEARNING_GRAPH.md` (or `.agents/LEARNING.md`) file. The agent gets smarter and more aligned with the user over time.
+
+### Trigger Conditions
+Activate this skill when:
+- The user corrects your code (e.g., "Don't use `any`, use `unknown`", or "Always use Arrow functions").
+- You just spent a long time fixing a very complex bug.
+- The user says "Remember this for the future."
+
+### Core Concepts
+
+#### 1. Memory Encoding
+Whenever a hard lesson is learned or a strong user preference is stated, do not just say "I will remember that." You must physically write it down.
+Use the `multi_replace_file_content` or `write_to_file` tool to append the lesson to the memory file.
+
+#### 2. Memory Format (`LEARNING_GRAPH.md`)
+```markdown
+# 🧠 Self-Evolving Memory Graph
+
+## User Stylistic Preferences
+- [2026-08-12] Always use `Tailwind v4 @theme` variables instead of arbitrary values like `text-[#ff0000]`.
+- [2026-08-12] Prefer functional components with `const` over `function` declarations.
+
+## Hard-Learned Lessons (Avoid these mistakes)
+- [2026-08-11] Supabase RLS policies: Never use `auth.uid()` in a `FOR ALL` policy without checking role claims. It caused an infinite recursion bug in `users` table.
+
+## Architectural Decisions
+- [2026-08-10] State Management: We exclusively use `Zustand` for global state. Do not suggest Redux.
+```
+
+#### 3. Memory Retrieval
+Ensure `session-context-loader` is configured to read this memory file at the start of every new chat session. 
+
+#### 4. Cross-Project Episodic Memory (Global Context)
+- **Continuous Capture**: For complex architectural decisions or bug fixes that span beyond a single project, extract the root cause and solution into a global knowledge graph (e.g., `~/.gemini/knowledge` or equivalent `<appDataDir>/knowledge` KI system).
+- **Contextual Retrieval**: On new sessions, search the global context directory for similar past issues before diving into complex bugs.
+- **Synthesis & Injection**: Inject the historical context directly into the current execution loop, stating explicitly: *"I remember we solved this architectural issue in Project X using technique Y. Applying the same proven pattern here."*
+
+---
+
+### Integration with Other Skills (MANDATORY)
+- `session-context-loader` — Essential for loading the `LEARNING_GRAPH.md` when a new session starts.
+- `project-context-mapper` — The context map tracks *where* things are; the memory graph tracks *how* and *why* things are built.
+
+### Referenced By Orchestrators (MANDATORY)
+- `brainstorming` — Add to "Discovery & Audit".
+- `zero-to-prod-orchestrator` — Phase 1 (Discovery & Architecture).
+
+---
+
+<a name="bahasa-indonesia"></a>
+## Bahasa Indonesia
+
+### Integrasi Orkestrasi
+Terhubung dan mengorkestrasi skill domain yang relevan seperti `brainstorming`, `zero-to-prod-orchestrator`, dan `project-context-mapper` untuk memastikan eksekusi yang kohesif.
+
+### Deskripsi
+Claude melupakan segalanya saat sesi chat baru dimulai. Skill ini memberi Anda "kekuatan super" berupa *Ingatan Jangka Panjang* yang berevolusi sendiri. Agen secara aktif mencatat preferensi koding pengguna, pelajaran dari *bug* yang sulit dipecahkan, dan keputusan arsitektur ke dalam file persisten `LEARNING_GRAPH.md`. 
+
+### Kondisi Pemicu
+- Saat pengguna mengoreksi gaya koding Anda (misal: "Gunakan bahasa Indonesia baku", "Jangan pakai `var`").
+- Setelah Anda berhasil memecahkan *bug* yang menghabiskan waktu lama.
+- Saat pengguna berkata "Ingat ini untuk proyek ke depannya."
+
+### Panduan Singkat
+- **Jangan Hanya Berjanji:** Jika pengguna mengoreksi Anda, jangan sekadar menjawab "Baik, saya akan ingat." LLM tidak punya ingatan bawaan antar-sesi. Anda **wajib** menulis koreksi tersebut ke dalam file `LEARNING_GRAPH.md` di folder `.agents/` atau direktori *root*.
+- **Kategorisasi Ingatan:** Pisahkan ingatan menjadi: Preferensi Gaya (*Style*), Pelajaran Berharga (*Lessons Learned*), dan Keputusan Arsitektur (*Decisions*).
+- **Semakin Lama Semakin Pintar:** Dengan membaca file ini di awal setiap percakapan, Anda tidak akan pernah mengulangi kesalahan yang sama dua kali.
+
+### Memori Episodik Lintas-Proyek (Konteks Global)
+- **Penangkapan Berkelanjutan**: Untuk bug yang rumit dan pola arsitektur tingkat lanjut, ekstrak solusi ke dalam grafik pengetahuan global (misal: `~/.gemini/knowledge`).
+- **Pencarian Kontekstual**: Sebelum memecahkan masalah kompleks baru, cari direktori konteks global untuk kasus serupa.
+- **Sintesis & Injeksi**: Secara proaktif sebutkan: *"Saya ingat kita menyelesaikan ini di Proyek X dengan cara Y, saya akan menerapkannya di sini."*

package/skills/self-healing-cloud-orchestrator/SKILL.md

@@ -1,57 +1,57 @@
----
-name: self-healing-cloud-orchestrator
-description: "Real-time log monitoring, crash detection, and auto-hotfixing code without human intervention / Pemantauan log real-time, deteksi kerusakan, dan perbaikan kode hotfix otomatis tanpa intervensi manusia."
-author: vibes-plug-swarm
----
-
-# Self-Healing Cloud Orchestrator (Code-to-Cloud Auto Remediation)
-
-[English](#english) | [Bahasa Indonesia](#bahasa-indonesia)
-
----
-
-<a name="english"></a>
-## English
-
-### Description
-Bridges the gap between code generation and Cloud Operations (CloudOps). This agent continuously monitors production or staging logs (via Vercel, AWS CloudWatch, Datadog, or Sentry MCP servers). When an anomaly, crash (OOM, Null Pointer), or configuration mismatch occurs, it autonomously downloads the stack trace, identifies the root cause in the source code, writes a patch, validates it locally, and pushes a hotfix deployment without requiring human intervention.
-
-### Trigger Conditions
-- Post-deployment in Phase 8, running as a persistent background daemon.
-- When an application goes down or experiences a spike in 5xx errors.
-- When `proactive-background-watcher` detects a critical health-check failure.
-
-### Operating Protocol
-1. **Telemetry Ingestion**: Listens to structured logs and APM alerts.
-2. **Root Cause Analysis (RCA)**: Parses stack traces and maps them to the local repository's current commit state.
-3. **Autonomous Patching**: Writes the necessary fix (e.g., adding missing env vars, fixing memory leaks, patching unhandled promise rejections).
-4. **Validation & Deploy**: Runs local unit/E2E tests via `autonomous-tdd-debugger` and commits the hotfix to trigger the CI/CD pipeline.
-
-## Orchestration & Integration
-- Utilizes `proactive-background-watcher` for continuous daemon execution.
-- Consumes data from `logging-error-tracking-expert` and `data-telemetry-expert`.
-- Delegates to `ci-cd-devops-architect` to push fixes and initiate deployments.
-
----
-
-<a name="bahasa-indonesia"></a>
-## Bahasa Indonesia
-
-### Deskripsi
-Menjembatani kesenjangan antara pembuatan kode dan Operasi Cloud (CloudOps). Agen ini secara terus-menerus memantau log produksi atau staging (melalui server MCP Vercel, AWS CloudWatch, Datadog, atau Sentry). Ketika terjadi anomali, kerusakan (OOM, Null Pointer), atau ketidaksesuaian konfigurasi, agen ini secara otonom mengunduh *stack trace*, mengidentifikasi akar masalah pada *source code*, menulis *patch*, memvalidasinya secara lokal, dan meluncurkan *deployment hotfix* tanpa intervensi manusia.
-
-### Kondisi Pemicu
-- Pasca-deployment di Fase 8, berjalan sebagai daemon latar belakang yang persisten.
-- Ketika sebuah aplikasi tumbang (*down*) atau mengalami lonjakan *error* 5xx.
-- Ketika `proactive-background-watcher` mendeteksi kegagalan pengecekan kesehatan (*health-check*) yang kritis.
-
-### Protokol Operasi
-1. **Penyerapan Telemetri**: Mendengarkan log terstruktur dan peringatan APM.
-2. **Analisis Akar Masalah (RCA)**: Menguraikan *stack trace* dan memetakannya ke status komit repositori lokal saat ini.
-3. **Patching Otonom**: Menulis perbaikan yang diperlukan (misal: menambahkan *environment variable* yang hilang, memperbaiki kebocoran memori, menambal *unhandled promise rejection*).
-4. **Validasi & Deploy**: Menjalankan *unit test/E2E test* lokal melalui `autonomous-tdd-debugger` lalu melakukan komit untuk memicu pipeline CI/CD.
-
-## Integrasi Orkestrasi
-- Menggunakan `proactive-background-watcher` untuk eksekusi daemon berkelanjutan.
-- Mengonsumsi data dari `logging-error-tracking-expert` dan `data-telemetry-expert`.
-- Mendelegasikan tugas ke `ci-cd-devops-architect` untuk mendorong perbaikan (*push*) dan memulai deployment.
+---
+name: self-healing-cloud-orchestrator
+description: "Real-time log monitoring, crash detection, and auto-hotfixing code without human intervention / Pemantauan log real-time, deteksi kerusakan, dan perbaikan kode hotfix otomatis tanpa intervensi manusia."
+author: "Roedy Rustam"
+---
+
+# Self-Healing Cloud Orchestrator (Code-to-Cloud Auto Remediation)
+
+[English](#english) | [Bahasa Indonesia](#bahasa-indonesia)
+
+---
+
+<a name="english"></a>
+## English
+
+### Description
+Bridges the gap between code generation and Cloud Operations (CloudOps). This agent continuously monitors production or staging logs (via Vercel, AWS CloudWatch, Datadog, or Sentry MCP servers). When an anomaly, crash (OOM, Null Pointer), or configuration mismatch occurs, it autonomously downloads the stack trace, identifies the root cause in the source code, writes a patch, validates it locally, and pushes a hotfix deployment without requiring human intervention.
+
+### Trigger Conditions
+- Post-deployment in Phase 8, running as a persistent background daemon.
+- When an application goes down or experiences a spike in 5xx errors.
+- When `proactive-background-watcher` detects a critical health-check failure.
+
+### Operating Protocol
+1. **Telemetry Ingestion**: Listens to structured logs and APM alerts.
+2. **Root Cause Analysis (RCA)**: Parses stack traces and maps them to the local repository's current commit state.
+3. **Autonomous Patching**: Writes the necessary fix (e.g., adding missing env vars, fixing memory leaks, patching unhandled promise rejections).
+4. **Validation & Deploy**: Runs local unit/E2E tests via `autonomous-tdd-debugger` and commits the hotfix to trigger the CI/CD pipeline.
+
+## Orchestration & Integration
+- Utilizes `proactive-background-watcher` for continuous daemon execution.
+- Consumes data from `logging-error-tracking-expert` and `data-telemetry-expert`.
+- Delegates to `ci-cd-devops-architect` to push fixes and initiate deployments.
+
+---
+
+<a name="bahasa-indonesia"></a>
+## Bahasa Indonesia
+
+### Deskripsi
+Menjembatani kesenjangan antara pembuatan kode dan Operasi Cloud (CloudOps). Agen ini secara terus-menerus memantau log produksi atau staging (melalui server MCP Vercel, AWS CloudWatch, Datadog, atau Sentry). Ketika terjadi anomali, kerusakan (OOM, Null Pointer), atau ketidaksesuaian konfigurasi, agen ini secara otonom mengunduh *stack trace*, mengidentifikasi akar masalah pada *source code*, menulis *patch*, memvalidasinya secara lokal, dan meluncurkan *deployment hotfix* tanpa intervensi manusia.
+
+### Kondisi Pemicu
+- Pasca-deployment di Fase 8, berjalan sebagai daemon latar belakang yang persisten.
+- Ketika sebuah aplikasi tumbang (*down*) atau mengalami lonjakan *error* 5xx.
+- Ketika `proactive-background-watcher` mendeteksi kegagalan pengecekan kesehatan (*health-check*) yang kritis.
+
+### Protokol Operasi
+1. **Penyerapan Telemetri**: Mendengarkan log terstruktur dan peringatan APM.
+2. **Analisis Akar Masalah (RCA)**: Menguraikan *stack trace* dan memetakannya ke status komit repositori lokal saat ini.
+3. **Patching Otonom**: Menulis perbaikan yang diperlukan (misal: menambahkan *environment variable* yang hilang, memperbaiki kebocoran memori, menambal *unhandled promise rejection*).
+4. **Validasi & Deploy**: Menjalankan *unit test/E2E test* lokal melalui `autonomous-tdd-debugger` lalu melakukan komit untuk memicu pipeline CI/CD.
+
+## Integrasi Orkestrasi
+- Menggunakan `proactive-background-watcher` untuk eksekusi daemon berkelanjutan.
+- Mengonsumsi data dari `logging-error-tracking-expert` dan `data-telemetry-expert`.
+- Mendelegasikan tugas ke `ci-cd-devops-architect` untuk mendorong perbaikan (*push*) dan memulai deployment.