Awesome Copilot Indonesia

Akselerasi SDLC Anda dengan Custom AI Agents

Koleksi agen, skill, aturan, dan prompt kustom yang siap pakai untuk mengoptimalkan alur kerja AI-assisted development Anda di berbagai platform (Antigravity, Copilot, OpenCode, dll).

Mulai Instalasi Jelajahi Agen
terminal 
$ awesome-copilot-id init
āœ” Platform detected: Google Antigravity
āœ” memory.instructions.md created
āœ” 10 custom agents installed successfully!

$ /sdlc-write-code [Bypass SDLC] implement auth
Analyzing codebase...
> Injecting auth middleware...
> All tests passed (12/12)

Overview

Awesome Copilot Indonesia adalah koleksi konfigurasi agen AI, skill, aturan, dan prompt kustom yang siap pakai untuk mempercepat siklus hidup pengembangan perangkat lunak (SDLC) Anda.

Custom Agents

Agen AI khusus untuk setiap fase pengembangan (PRD, Technical Specs, Strategic Planning, Coding, hingga Review).

Specialized Skills

Kemampuan khusus seperti Architecture Mapper (/sdlc-map-architecture), Karpathy Guidelines, Polya Heuristic Coder, Ponytail lazy dev, TDD Implement, dan Fable Protocol untuk eksekusi tugas otonom yang kompleks.

Rules & Instructions

Panduan coding terbaik yang disesuaikan dengan bahasa pemrograman dan framework populer untuk menjaga kualitas kode Anda.

BYOK Copilot Config

Template konfigurasi siap pakai untuk menggunakan API key Anda sendiri (OpenRouter, DeepSeek, dll) di VS Code Copilot Chat.

Dukungan Multi-Platform

Konfigurasi siap pakai untuk Google Antigravity (`.agents`), Claude Code (`.claude`), OpenCode, Copilot, CommandCode, ChatGPT Codex, Pi Dev, dan Oh My Pi (`omp`).

Getting Started

Pilih metode instalasi otomatis yang cepat atau lakukan instalasi manual sesuai dengan platform asisten AI yang Anda gunakan.

Prasyarat Platform

  • Google Antigravity
  • GitHub Copilot
  • OpenCode
  • CommandCode
  • ChatGPT Codex
  • Claude Code
  • Pi Dev
  • Oh My Pi (omp)

Instal bootstrapper skill langsung menggunakan CLI npx skills:

Standard SDLC 12 Phases

1. Terminal:

npx skills add GulajavaMinistudio/awesome-copilot-id/.agents/skills/sdlc-init

2. AI Chat:

/sdlc-init setup this project
TDD-Spec SDLC 21 Skills

1. Terminal:

npx skills add GulajavaMinistudio/awesome-copilot-id/tdd-spec-skills/.agents/skills/tdd-init

2. AI Chat:

/tdd-init setup this project
PĆ³lya Heuristic Coder 11 Skills

1. Terminal:

npx skills add GulajavaMinistudio/awesome-copilot-id/polya-coder/.agents/skills/polya-init

2. AI Chat:

/polya-init setup this project

3. AI Chat (usage):

/polya-router
PĆ³lya SDLC Suite 14 Skills

1. Terminal:

npx skills add GulajavaMinistudio/awesome-copilot-id/polya-sdlc/.agents/skills/to-init

2. AI Chat:

/to-init setup this project

3. AI Chat (usage):

/to-ask-help
Otonom & Aman: Agen AI akan mengunduh AGENTS.md dan folder .agents/ secara otonom, menggabungkan aturan global, serta mengamankan catatan memory.instructions.md Anda secara otomatis.

Jalankan satu perintah (one-liner) di terminal pada direktori root proyek Anda:

Linux / macOS (Bash/Zsh)

curl -fsSL https://raw.githubusercontent.com/GulajavaMinistudio/awesome-copilot-id/main/install.sh | bash

Windows Terminal (PowerShell)

irm https://raw.githubusercontent.com/GulajavaMinistudio/awesome-copilot-id/main/install.ps1 | iex
Penting: Setelah instalasi selesai, buka file AGENTS.md di root proyek Anda dan ubah:
1. Placeholder [Your Application Name] dengan nama proyek Anda.
2. Tambahkan [Project Description] yang singkat.
3. Sesuaikan preferensi Language: sesuai kebutuhan Anda.
  1. Clone repositori ini: 
    git clone https://github.com/GulajavaMinistudio/awesome-copilot-id.git
  2. Pilih konfigurasi platform Anda:

    Salin folder platform yang relevan ke root proyek Anda. Contoh untuk Google Antigravity, salin folder .agents. Untuk GitHub Copilot, salin folder .github.

  3. Salin file AGENTS.md ke root proyek: 
    cp awesome-copilot-id/AGENTS.md ./

    Penting: Setelah disalin, buka AGENTS.md dan ubah:
    1. Placeholder [Your Application Name] dengan nama proyek Anda.
    2. Tambahkan [Project Description] yang singkat.
    3. Sesuaikan preferensi Language: sesuai kebutuhan Anda.

  4. Mulai ulang IDE atau Asisten AI Anda untuk menerapkan perubahan.
13 Agen Spesialis

Custom Agents

Gunakan agen AI kustom yang berspesialisasi dalam peran dan tugas tertentu dalam siklus pengembangan.

Phase 1-3: Strategy & Specs
Phase 4: Execution & Review
Phase 5: Documentation
Supplementary (Bypass)
Pemeta Arsitektur (Pre-Phase)

/sdlc-map-architecture

Pre-Phase: Architecture Mapping

Peran & Tanggung Jawab Utama:

Memindai dan memetakan topologi repositori, struktur direktori, titik masuk, dan konfigurasi runtime ke dalam docs/ARCHITECTURE.md.

Paling Cocok Untuk:

Mendokumentasikan topologi sistem, memetakan basis kode baru/legacy, dan menjaga docs/ARCHITECTURE.md selalu mutakhir.

Contoh Prompt Penggunaan:

/sdlc-map-architecture map repository architecture to docs/ARCHITECTURE.md
Discovery Explorer (Phase 0)

/sdlc-explore-ideas

Phase 0: Discovery

Peran & Tanggung Jawab Utama:

Mengeksplorasi basis kode secara sistematis, memberikan kritik arsitektur, mengidentifikasi technical debt, dan menghasilkan draft penemuan proyek awal (Phase 0 Discovery).

Paling Cocok Untuk:

Memahami proyek baru atau lama, menganalisis utang teknis, dan menghasilkan peta arsitektur sebelum merumuskan PRD.

Contoh Prompt Penggunaan:

/sdlc-explore-ideas explore the codebase and write a discovery draft for the new shopping cart
Product Requirements (Phase 1)

/sdlc-draft-prd

Phase 1: PRD

Peran & Tanggung Jawab Utama:

Menghasilkan Product Requirements Document (PRD) yang komprehensif, mencakup User Stories, alur pengguna, dan kriteria penerimaan.

Paling Cocok Untuk:

Menulis dokumentasi kebutuhan bisnis dan fungsionalitas fitur baru sebelum spesifikasi teknis ditulis.

Contoh Prompt Penggunaan:

/sdlc-draft-prd create a PRD for the user authentication module based on @discovery-draft.md
Clarification Gatekeeper (Post-PRD)

/sdlc-clarify-reqs

Quality Gate 1

Peran & Tanggung Jawab Utama:

Melakukan interogasi terhadap PRD, Spesifikasi Teknis, atau Rencana Implementasi untuk menemukan ambiguitas, asumsi tersembunyi, dan edge-case yang terlewat.

Paling Cocok Untuk:

Pengecekan kualitas kebutuhan sebelum melangkah ke spesifikasi teknis atau penulisan kode.

Contoh Prompt Penggunaan:

/sdlc-clarify-reqs check the requirements in @auth-prd.md and highlight potential edge cases
Technical Architect (Phase 2)

/sdlc-define-specs

Phase 2: Spec

Peran & Tanggung Jawab Utama:

Membuat spesifikasi teknis detail (kontrak API, skema DB, model data) dan mendokumentasikan keputusan arsitektur penting (ADR).

Paling Cocok Untuk:

Merancang cetak biru teknis sistem yang terstruktur sebelum mulai menulis kode.

Contoh Prompt Penggunaan:

/sdlc-define-specs create a technical specification for the cart checkout based on @cart-prd.md
Traceability & Consistency Auditor

/sdlc-audit-consistency

Audit & Quality Gate

Peran & Tanggung Jawab Utama:

Mengaudit konsistensi dan ketertelusuran dokumen (PRD vs Spec vs Plan) untuk mendeteksi hilangnya cakupan kebutuhan dan scope creep.

Paling Cocok Untuk:

Memastikan alur SDLC konsisten dan tidak ada fitur yang terlupa atau melenceng dari rencana awal.

Contoh Prompt Penggunaan:

/sdlc-audit-consistency check consistency between @cart-prd.md and @cart-spec.md
Clarification Gatekeeper (Post-Spec)

/sdlc-clarify-reqs

Quality Gate 2

Peran & Tanggung Jawab Utama:

Melakukan interogasi terhadap PRD, Spesifikasi Teknis, atau Rencana Implementasi untuk menemukan ambiguitas, asumsi tersembunyi, dan edge-case yang terlewat.

Paling Cocok Untuk:

Pengecekan kualitas kebutuhan sebelum melangkah ke spesifikasi teknis atau penulisan kode.

Contoh Prompt Penggunaan:

/sdlc-clarify-reqs check if there are missing technical constraints in @cart-spec.md
Implementation Planner (Phase 3)

/sdlc-plan-tasks

Phase 3: Plan

Peran & Tanggung Jawab Utama:

Memecah Spesifikasi Teknis menjadi langkah-langkah implementasi strategis yang terstruktur dan dapat dieksekusi secara berurutan.

Paling Cocok Untuk:

Menyusun jadwal tugas coding yang terperinci dan aman sebelum developer mulai menulis kode.

Contoh Prompt Penggunaan:

/sdlc-plan-tasks create a step-by-step implementation plan for @cart-spec.md
Clarification Gatekeeper (Post-Plan)

/sdlc-clarify-reqs

Quality Gate 3

Peran & Tanggung Jawab Utama:

Melakukan interogasi terhadap PRD, Spesifikasi Teknis, atau Rencana Implementasi untuk menemukan ambiguitas, asumsi tersembunyi, dan edge-case yang terlewat.

Paling Cocok Untuk:

Pengecekan kualitas kebutuhan sebelum melangkah ke spesifikasi teknis atau penulisan kode.

Contoh Prompt Penggunaan:

/sdlc-clarify-reqs verify if @cart-plan.md is missing rollback strategies
Autonomous Developer (Phase 4)

/sdlc-write-code

Phase 4: Execution

Peran & Tanggung Jawab Utama:

Agen developer otonom berkinerja tinggi yang menulis fitur, melakukan refaktorisasi bedah, dan menulis tes unit/widget yang kokoh.

Paling Cocok Untuk:

Menulis kode fitur, merefaktor modul, memperbaiki bug, dan menjalankan tes langsung di ruang kerja.

Contoh Prompt Penggunaan:

/sdlc-write-code execute task 1 from @cart-plan.md
5-Axis Code & Security Reviewer

/sdlc-code-review

Review & Security

Peran & Tanggung Jawab Utama:

Meninjau kualitas kode berdasarkan prinsip SOLID, Clean Code, dan praktik terbaik keamanan, serta menghasilkan rencana refaktorisasi terstruktur.

Paling Cocok Untuk:

Menjaga kualitas kode dan standar keamanan sebelum menggabungkan perubahan ke branch utama.

Contoh Prompt Penggunaan:

/sdlc-code-review review the recent changes in @cart.service.ts against SOLID principles
Root Cause & Bug Remediation

/sdlc-bug-report

RCA Diagnosis

Peran & Tanggung Jawab Utama:

Menganalisis laporan bug, melakukan Analisis Akar Penyebab (RCA), dan menghasilkan rencana perbaikan bug bedah dengan strategi rollback yang aman.

Paling Cocok Untuk:

Menyelesaikan bug kompleks di codebase tanpa menimbulkan efek samping atau regresi.

Contoh Prompt Penggunaan:

/sdlc-bug-report analyze the crash report in @crash-log.txt
Diataxis Documentation (Phase 5)

/sdlc-generate-docs

Phase 5: Docs

Peran & Tanggung Jawab Utama:

Spesialis dokumentasi teknis yang menulis panduan pengguna, panduan API, dan dokumen referensi yang jelas dengan kerangka kerja Diataxis.

Paling Cocok Untuk:

Menulis panduan pengguna, dokumentasi API, dan penjelasan sistem yang mudah dipahami.

Contoh Prompt Penggunaan:

/sdlc-generate-docs write an API reference guide based on @spec-shopping-cart.md
Rapid Fixer & Refactor (Bypass)

/code-janitor

Rapid Bypass

Peran & Tanggung Jawab Utama:

Agen minimalis super cepat yang menggabungkan perencanaan mikro dan eksekusi dalam satu alur kerja untuk perbaikan ad-hoc dan refaktor kecil di luar proses SDLC yang ketat.

Paling Cocok Untuk:

Perbaikan bug cepat, pembersihan kode, dan refactoring minor yang tidak membutuhkan dokumen PRD-Spec-Plan lengkap.

Contoh Prompt Penggunaan:

/code-janitor tolong refaktor fungsi calculateTotal di @cart.js agar lebih ringkas
Interactive AI Mentor (Bypass)

/guided-learning

AI Mentor

Peran & Tanggung Jawab Utama:

AI-Guided Learning mentor. Memandu pengguna selangkah demi selangkah untuk membangun proyek atau mengimplementasikan rencana tanpa menulis kode secara langsung.

Paling Cocok Untuk:

Digabungkan bersama /sdlc-write-code atau /code-janitor sebagai alternatif "Mentor Mode" agar pengguna dapat belajar mengetik kodenya sendiri.

Contoh Prompt Penggunaan:

/sdlc-write-code implement @plan.md using guided-learning

Memindai dan memetakan topologi repositori, struktur direktori, titik masuk, dan konfigurasi runtime ke dalam docs/ARCHITECTURE.md.

Paling Cocok Untuk:

Mendokumentasikan topologi sistem, memetakan basis kode baru/legacy, dan menjaga docs/ARCHITECTURE.md selalu mutakhir.

Contoh Prompt Penggunaan:

/sdlc-map-architecture map repository architecture to docs/ARCHITECTURE.md

Mengeksplorasi basis kode secara sistematis, memberikan kritik arsitektur, mengidentifikasi technical debt, dan menghasilkan draft penemuan proyek awal (Phase 0 Discovery).

Paling Cocok Untuk:

Memahami proyek baru atau lama, menganalisis utang teknis, dan menghasilkan peta arsitektur sebelum merumuskan PRD.

Contoh Prompt Penggunaan:

/sdlc-explore-ideas explore the codebase and write a discovery draft for the new shopping cart

Menghasilkan Product Requirements Document (PRD) yang komprehensif, mencakup User Stories, alur pengguna, dan kriteria penerimaan.

Paling Cocok Untuk:

Menulis dokumentasi kebutuhan bisnis dan fungsionalitas fitur baru sebelum spesifikasi teknis ditulis.

Contoh Prompt Penggunaan:

/sdlc-draft-prd create a PRD for the user authentication module based on @discovery-draft.md

Melakukan interogasi terhadap PRD, Spesifikasi Teknis, atau Rencana Implementasi untuk menemukan ambiguitas, asumsi tersembunyi, dan edge-case yang terlewat.

Paling Cocok Untuk:

Pengecekan kualitas kebutuhan sebelum melangkah ke spesifikasi teknis atau penulisan kode.

Contoh Prompt Penggunaan:

/sdlc-clarify-reqs check the requirements in @auth-prd.md and highlight potential edge cases

Membuat spesifikasi teknis detail (kontrak API, skema DB, model data) dan mendokumentasikan keputusan arsitektur penting (ADR).

Paling Cocok Untuk:

Merancang cetak biru teknis sistem yang terstruktur sebelum mulai menulis kode.

Contoh Prompt Penggunaan:

/sdlc-define-specs create a technical specification for the cart checkout based on @cart-prd.md

Mengaudit konsistensi dan ketertelusuran dokumen (PRD vs Spec vs Plan) untuk mendeteksi hilangnya cakupan kebutuhan dan scope creep.

Paling Cocok Untuk:

Memastikan alur SDLC konsisten dan tidak ada fitur yang terlupa atau melenceng dari rencana awal.

Contoh Prompt Penggunaan:

/sdlc-audit-consistency check consistency between @cart-prd.md and @cart-spec.md

Melakukan interogasi terhadap PRD, Spesifikasi Teknis, atau Rencana Implementasi untuk menemukan ambiguitas, asumsi tersembunyi, dan edge-case yang terlewat.

Paling Cocok Untuk:

Pengecekan kualitas kebutuhan sebelum melangkah ke spesifikasi teknis atau penulisan kode.

Contoh Prompt Penggunaan:

/sdlc-clarify-reqs check if there are missing technical constraints in @cart-spec.md

Memecah Spesifikasi Teknis menjadi langkah-langkah implementasi strategis yang terstruktur dan dapat dieksekusi secara berurutan.

Paling Cocok Untuk:

Menyusun jadwal tugas coding yang terperinci dan aman sebelum developer mulai menulis kode.

Contoh Prompt Penggunaan:

/sdlc-plan-tasks create a step-by-step implementation plan for @cart-spec.md

Melakukan interogasi terhadap PRD, Spesifikasi Teknis, atau Rencana Implementasi untuk menemukan ambiguitas, asumsi tersembunyi, dan edge-case yang terlewat.

Paling Cocok Untuk:

Pengecekan kualitas kebutuhan sebelum melangkah ke spesifikasi teknis atau penulisan kode.

Contoh Prompt Penggunaan:

/sdlc-clarify-reqs verify if @cart-plan.md is missing rollback strategies

Agen developer otonom berkinerja tinggi yang menulis fitur, melakukan refaktorisasi bedah, dan menulis tes unit/widget yang kokoh.

Paling Cocok Untuk:

Menulis kode fitur, merefaktor modul, memperbaiki bug, dan menjalankan tes langsung di ruang kerja.

Contoh Prompt Penggunaan:

/sdlc-write-code execute task 1 from @cart-plan.md

Meninjau kualitas kode berdasarkan prinsip SOLID, Clean Code, dan praktik terbaik keamanan, serta menghasilkan rencana refaktorisasi terstruktur.

Paling Cocok Untuk:

Menjaga kualitas kode dan standar keamanan sebelum menggabungkan perubahan ke branch utama.

Contoh Prompt Penggunaan:

/sdlc-code-review review the recent changes in @cart.service.ts against SOLID principles

Menganalisis laporan bug, melakukan Analisis Akar Penyebab (RCA), dan menghasilkan rencana perbaikan bug bedah dengan strategi rollback yang aman.

Paling Cocok Untuk:

Menyelesaikan bug kompleks di codebase tanpa menimbulkan efek samping atau regresi.

Contoh Prompt Penggunaan:

/sdlc-bug-report analyze the crash report in @crash-log.txt

Spesialis dokumentasi teknis yang menulis panduan pengguna, panduan API, dan dokumen referensi yang jelas dengan kerangka kerja Diataxis.

Paling Cocok Untuk:

Menulis panduan pengguna, dokumentasi API, dan penjelasan sistem yang mudah dipahami.

Contoh Prompt Penggunaan:

/sdlc-generate-docs write an API reference guide based on @spec-shopping-cart.md

Agen minimalis super cepat yang menggabungkan perencanaan mikro dan eksekusi dalam satu alur kerja untuk perbaikan ad-hoc dan refaktor kecil di luar proses SDLC yang ketat.

Paling Cocok Untuk:

Perbaikan bug cepat, pembersihan kode, dan refactoring minor yang tidak membutuhkan dokumen PRD-Spec-Plan lengkap.

Contoh Prompt Penggunaan:

/code-janitor tolong refaktor fungsi calculateTotal di @cart.js agar lebih ringkas

AI-Guided Learning mentor. Memandu pengguna selangkah demi selangkah untuk membangun proyek atau mengimplementasikan rencana tanpa menulis kode secara langsung.

Paling Cocok Untuk:

Digabungkan bersama /sdlc-write-code atau /code-janitor sebagai alternatif "Mentor Mode" agar pengguna dapat belajar mengetik kodenya sendiri.

Contoh Prompt Penggunaan:

/sdlc-write-code implement @plan.md using guided-learning
Test-Driven Development

TDD-Spec SDLC Package (21 Specialized Skills)

A dedicated test-first engineering suite combining pure Test-Driven Development (TDD), Behavior-Driven Development (BDD), and Executable Specifications.

Featured Navigator

AI Guide & Interactive Navigator (/tdd-ask-help)

Never wonder what to do next. /tdd-ask-help scans your project workspace artifacts in real-time and recommends the exact next slash command with ready-to-use prompts.

/tdd-ask-help /tdd-ask-help [pertanyaan Anda]
tdd-navigator 
Q: "Where is my project right now and what is next?"
Diagnosis: Found approved PRD at docs/prd/auth.md
Next Step: /tdd-spec
Why: Blueprint & Pre-Agreed Test Seams required before planning.

Q: "Found 500 error on checkout coupon, how to fix?"
Diagnosis: Functional bug in cart calculator.
Next Step: /tdd-bug-report (Prove-It failing test first)
/tdd-init Bootstrap

TDD Bootstrapper

Bootstraps and calibrates project governance and auto-maps existing codebases.

/tdd-init setup this project
/tdd-map-architecture Map

Architecture Mapper

Scans directory topology, test runners, and public test seams into docs/ARCHITECTURE.md.

/tdd-map-architecture map codebase
/tdd-explore-ideas Discovery

Idea Explorer

5-Step Idea Assessment with measurable hypothesis test metrics.

/tdd-explore-ideas explore @brief.md
/tdd-prd PRD

TDD Product Manager

Authored BDD user stories with executable Given-When-Then acceptance scenarios.

/tdd-prd draft PRD from @discovery.md
/tdd-clarify Quality Gate

Requirements Clarifier

Doubt-Driven interrogator uncovering ambiguities with 40/30/30 Readiness Scoring.

/tdd-clarify interrogate @prd.md
/tdd-analyze Impact Audit

Codebase Analyst

Pre-execution impact analysis mapping blast radius and mocking traps.

/tdd-analyze audit blast radius @spec.md
/tdd-spec Blueprint

Spec Architect

Technical blueprint with interface contracts and Pre-Agreed Test Seams.

/tdd-spec define spec from @prd.md
/tdd-checklist Matrix

Checklist Generator

Translates specifications into a Test-Case Inventory Matrix.

/tdd-checklist generate matrix @spec.md
/tdd-plan-tasks Plan

Tracer-Bullet Planner

Breaks specifications into vertical slices with explicit RED-GREEN-VERIFY steps.

/tdd-plan-tasks plan from @spec.md
/tdd-write-code Execution

TDD Code Engine

Strict Red-Green-Refactor implementation engine with Floor-Guard anti-cheat.

/tdd-write-code execute @plan.md
/tdd-code-review 5-Axis

5-Axis Reviewer

5-Axis code & test review (Spec, Health, Security, Perf, Test Efficacy).

/tdd-code-review review @spec.md
/tdd-bug-report Prove-It

Prove-It Bug Diagnostician

RCA with Prove-It Pattern: mandates writing a failing regression test first.

/tdd-bug-report diagnose @error.log
/tdd-generate-docs DiƔtaxis

DiƔtaxis Docs Writer

User documentation using passing test suites as Living Examples.

/tdd-generate-docs generate docs
/tdd-retro Retro

Retrospective Optimizer

Optimizes test suite run speeds, flakiness, and synchronizes memory.

/tdd-retro analyze sprint
/tdd-ask-help Navigator

TDD SDLC Navigator

Interactive AI project navigator diagnosing project phase and routing scenarios.

/tdd-ask-help where are we now?
/tdd-pair-coach Coach

Pair Coach

Interactive Socratic AI pair programming mentor guiding developers step-by-step.

/tdd-pair-coach guide me on @task-1.md
/tdd-mutation-test MSI Audit

Mutation Auditor

Mutation testing runner auditing assertion efficacy (MSI >= 80%).

/tdd-mutation-test audit @cart.spec.js
/tdd-configure-ci CI Pipeline

CI Gatekeeper

Configures CI/CD pipelines to mechanically enforce CONSTRAINTS.md floor-guards.

/tdd-configure-ci setup github actions
/tdd-generate-fixtures Fixtures

Fixture Architect

Generates type-safe test factories, seeders, and fixtures from spec models.

/tdd-generate-fixtures for @spec.md
/tdd-refactor-legacy Golden Master

Legacy Modernizer

Pins untested legacy code behavior with Golden Master characterization tests.

/tdd-refactor-legacy pin @legacy.js
memory-manager Memory

Context Persistence

Persists session checkpoints and maintains knowledge in memory.instructions.md.

save session progress to memory
1945 Mathematical Heuristics & Clean Architecture

PĆ³lya Heuristic Coder (Modular Engineering Suite)

A disciplined problem-solving persona combining George PĆ³lya's 1945 mathematical heuristic methodology with Uncle Bob's Clean Code, Clean Architecture, and SOLID principles.

Tentang PĆ³lya Edisi SDLC

PĆ³lya SDLC adalah varian mandiri yang dioptimalkan khusus untuk alur SDLC lengkap dengan alias perintah berbahasa Indonesia (/to-*). Dilengkapi pembuat inisialisasi proyek sendiri (to-init), router utama (/to-ask-help), dan penegakan Clean Architecture ketat di setiap 4 tahap pemecahan masalah George PĆ³lya.

Master SDLC Router

Master SDLC Navigator (/to-ask-help)

Unified interactive router for the PĆ³lya SDLC workflow. It analyzes project context, diagnoses current phase, and recommends the exact /to-* slash command to run next.

/to-ask-help /to-init /to-explore /to-prd /to-spec /to-plan /to-code
polya-sdlc-engine 
Prompt: "I need to implement multi-tenant authentication."
Analysis: Greenfield domain boundary; requires PRD and specification first.
PĆ³lya Stage: Stage 1 (Understand Problem) -> /to-prd
Quality Gate: Pause Rule enforced: Plan approval mandatory before coding.
1

Stage 1: Understand the Problem

Analyze unknowns, verify condition data, explore existing codebase topography, and author definitive specifications with ambiguity elimination.

/to-ask-help Triage
Triage: Master SDLC Navigator

Master interactive router that diagnoses project phase, analyzes current artifacts, and suggests the next /to-* step.

/to-ask-help
/to-init Setup
Setup: PĆ³lya SDLC Scaffolding

Interactive project initializer that sets up directory structures, AGENTS.md, rules, and skills for PĆ³lya SDLC.

/to-init setup this project
/to-explore Phase 0
Phase 0: Discovery & Topography

Phase 0 discovery and exploration. Analyzes requirements, critiques architecture, and prepares project discovery draft.

/to-explore @brief.md
/to-prd Phase 1
Phase 1: Product Requirements

Drafts structured Product Requirements Document (PRD) with user stories, acceptance criteria, and metric definitions.

/to-prd @docs/discovery.md
/to-spec Phase 2
Phase 2: Technical Specification

Defines technical specification with data contracts, architecture seams, API designs, and DB schemas.

/to-spec @docs/prd/checkout-prd.md
/to-clarify Checkpoint
Checkpoint: Clarification & Audit

Interrogates PRD, Spec, or Plan to expose ambiguities, edge cases, hidden assumptions, and calculate readiness score.

/to-clarify @docs/spec/checkout-spec.md
2

Stage 2: Devise a Plan (The Pause Rule)

Conceive a step-by-step strategy, slice into vertical tracer bullets, and enforce the mandatory Pause Rule to ensure user approval before execution.

/to-plan Phase 3
Phase 3: Tracer-Bullet Planning

Creates actionable vertical-slice implementation plans (Tracer Bullets) strictly derived from technical specs.

/to-plan @docs/spec/checkout-spec.md
/to-map Topography
Topography: Living Architecture Map

Maps repository architecture topography, module boundaries, and keeps docs/ARCHITECTURE.md evergreen.

/to-map scan repository topography
/to-memory Context
Context: Persistent Memory Manager

Persistent memory manager that records session progress, context, and decisions to memory.instructions.md.

/to-memory save active session context
3

Stage 3: Carry Out the Plan

Implement code incrementally with Micro-TDD discipline, maintaining single-responsibility and dependency inversion at every commit.

/to-code Phase 4
Phase 4: Clean Micro-TDD Coding

Senior software engineer executing vertical slices with Micro-TDD discipline, Pause Rules, and zero test suppression.

/to-code implement @docs/plan/checkout-plan.md
/to-quick-fix Fast-Track
Fast-Track: Surgical Fixer (Janitor)

Fast-track code fixer combining planning and execution for minor bug fixes, cleanups, and ad-hoc adjustments.

/to-quick-fix fix null pointer in OrderService.ts
4

Stage 4: Look Back (Review & Refactor)

Examine the result obtained, verify clean architecture integrity, conduct surgical root-cause analysis on bugs, and map living topography.

/to-review Phase 5
Phase 5: Clean Code & SOLID Review

Rigorous code review against SOLID, Clean Architecture, and technical spec compliance with actionable remediation.

/to-review @docs/spec/checkout-spec.md @docs/plan/checkout-plan.md
/to-docs Phase 6
Phase 6: Technical Documentation (DiƔtaxis)

Generates user and technical documentation structured under the DiƔtaxis framework (tutorials, how-to, reference, explanation).

/to-docs generate diataxis docs for checkout module
/to-fix Remediation
Remediation: Surgical Bug RCA

Performs surgical root-cause analysis (RCA) on bug reports, failure traces, and drafts phased remediation plans.

/to-fix analyze @bug-report.md
Context Governance

Context Injection Protocol (PĆ³lya SDLC Edition)

To prevent context loss, hallucinations, and to enforce strict traceability, you MUST attach or reference mandatory upstream documents in prompt context when invoking /to-* commands.

Action / Sub-Command Mandatory Upstream Document(s)
/to-ask-helpUser Brief / Feature request / Question
/to-initEmpty or existing repository root
/to-exploreSource code repository / Build configs
/to-prdProject Discovery Draft OR feature brief
/to-specApproved PRD OR comprehensive feature brief
/to-clarifyDiscovery Draft, Spec, OR Implementation Plan
/to-planApproved Technical Specification
/to-mapSource code repository / build configuration
/to-memoryActive session context / project status
/to-codeApproved Implementation Plan with Pause Rule clearance
/to-quick-fixSpecific source code file(s) to modify
/to-reviewTechnical Spec AND Implementation Plan
/to-docsTechnical Spec, Implementation Plan, OR Source Code files
/to-fixBug report, error log, OR failing test trace

*Note: For minor fixes and rapid refactoring, use /to-quick-fix to fast-track without full formal SDLC ceremony while maintaining surgical code quality.

Modular Triage Engine

Interactive Problem-Solving Triage (/polya-router)

A single master entrypoint that auto-detects existing repository topography, identifies project state, and routes you through George PĆ³lya's 4 stages with surgical precision.

/polya-router /polya-explore /polya-spec /polya-plan /polya-code
polya-heuristic-engine 
Prompt: "I need to implement payment gateway integration."
Analysis: Greenfield feature seam; requires Clean Architecture boundary.
PĆ³lya Stage: Stage 1 (Understand the Problem) -> /polya-spec
Quality Gate: Strict Pause Rule enforced before code generation.
1

Stage 1: Understand the Problem

Analyze unknowns, verify condition data, explore existing codebase topography, and author definitive specifications with ambiguity elimination.

/polya-explore Phase 0
Phase 0: Discovery & Topography

Explores existing codebases, critiques architecture, maps technical debt, and drafts raw discovery briefs.

/polya-explore @brief.md
/polya-spec Phase 1
Phase 1: Deep Specification

Designs Clean Architecture contracts, DDL schemas, and domain seams before touching any source code.

/polya-spec @brief.md
/polya-clarify Checkpoint
Checkpoint: Clarification Gate

Hunts down hidden assumptions, interrogates ambiguous requirements, and validates edge cases.

/polya-clarify @docs/spec/checkout-spec.md
2

Stage 2: Devise a Plan (The Pause Rule)

Conceive a step-by-step strategy, slice into vertical tracer bullets, and enforce the mandatory Pause Rule to ensure user approval before execution.

/polya-plan Phase 2
Phase 2: Tracer-Bullet Planning

Sequences vertical implementation tasks with explicit verification steps and strict Pause Rule compliance.

/polya-plan @docs/spec/checkout-spec.md
/polya-map Topography
Topography: Living Architecture Map

Maps repository topography, directory purposes, and runtime contracts to keep docs/ARCHITECTURE.md evergreen.

/polya-map
3

Stage 3: Carry Out the Plan

Implement code incrementally with Micro-TDD discipline, maintaining single-responsibility and dependency inversion at every commit.

/polya-code Phase 3
Phase 3: Clean Implementation

Writes surgical, production-ready code with unit/integration tests following Red-Green-Refactor and zero mock pollution.

/polya-code @docs/plan/checkout-plan.md
/polya-fast-track Bypass
Fast-Track Bypass (Ad-Hoc)

Fast-tracks minor fixes, typographical adjustments, and cleanups without requiring full SDLC paperwork.

/polya-fast-track fix button padding in @button.css
4

Stage 4: Look Back (Review & Refactor)

Examine the result obtained, verify clean architecture integrity, conduct surgical root-cause analysis on bugs, and map living topography.

/polya-review Phase 4
Phase 4: Clean Code & SOLID Review

Audits implementation against SOLID principles, detects code smells, and produces structured refactoring plans.

/polya-review @service.ts against @docs/spec/checkout-spec.md
/polya-docs Phase 6
Phase 6: Technical Documentation (DiƔtaxis)

Authors user and developer documentation partitioned across the 4 DiƔtaxis quadrants: Tutorials, How-To, Reference, and Explanation.

/polya-docs @docs/spec/checkout-spec.md
/polya-fix Phase 5
Phase 5: Surgical Bug Remediation

Performs medical-grade diagnosis, traces root causes, and implements targeted fixes with safe rollback strategies.

/polya-fix analyze @bug-report.md
Context Governance

Context Injection Protocol (PĆ³lya Heuristic Coder)

To prevent context loss, hallucinations, and to enforce strict traceability, you MUST attach or reference mandatory upstream documents in prompt context.

Action / Sub-Command Mandatory Upstream Document(s)
/polya-routerUser Brief / Feature request / Question
/polya-exploreSource code repository / Build configs
/polya-specDiscovery Draft OR comprehensive feature brief
/polya-clarifyDiscovery Draft, Spec, OR Implementation Plan
/polya-planApproved Technical Specification
/polya-codeApproved Implementation Plan with Pause Rule clearance
/polya-reviewTechnical Spec AND Implementation Plan
/polya-docsTechnical Spec, Implementation Plan, OR Source Code files
/polya-fixBug report, error log, OR failing test trace
/polya-fast-trackSpecific source code file(s) to modify
/polya-mapSource code repository / build configuration

*Note: For minor fixes and refactors, use /polya-fast-track to bypass the full formal SDLC ceremony while maintaining surgical code quality.

Interactive Workflows

Workflow & Examples

Pahami bagaimana siklus hidup pengembangan perangkat lunak (SDLC) dijalankan menggunakan rangkaian agen AI secara sekuensial.

Sekuens Alur SDLC Agen AI

Phase 0: Project Discovery (Penemuan Proyek)

Eksplorasi kode yang ada, analisis arsitektur, dan buat draf penemuan proyek awal sebelum merancang PRD.

Agent: /sdlc-explore-ideas Dokumen Input: Business Brief / Ideas 
/sdlc-explore-ideas explore the codebase and write a discovery draft for the new shopping cart feature based on @business-brief.md

Phase 1: Requirements & Clarification (Kebutuhan & Klarifikasi)

Buat PRD yang terperinci dan lakukan klarifikasi/interogasi mendalam untuk menghilangkan ambiguitas kebutuhan fitur.

Agent: /sdlc-draft-prd, /sdlc-clarify-reqs Dokumen Input: Project Discovery Draft 
/sdlc-draft-prd create a PRD for the shopping cart feature based on @discovery-draft.md
/sdlc-clarify-reqs interrogate the new @prd-shopping-cart.md for missing edge cases

Phase 2: Specification & Consistency Audit (Spesifikasi & Audit)

Tulis rancangan teknis (API, Skema DB) dan audit ketertelusurannya terhadap PRD untuk mencegah scope creep.

Agent: /sdlc-define-specs, /sdlc-audit-consistency Dokumen Input: Approved PRD 
/sdlc-define-specs design a technical specification based on @prd-shopping-cart.md
/sdlc-audit-consistency verify that @spec-shopping-cart.md strictly follows @prd-shopping-cart.md

Phase 3: Implementation Planning (Perencanaan Implementasi)

Pecah spesifikasi menjadi tiket vertikal (Tracer Bullets) yang dapat diuji secara independen.

Agent: /sdlc-plan-tasks Dokumen Input: Approved Technical Spec 
/sdlc-plan-tasks create a step-by-step implementation plan based on @spec-shopping-cart.md

Phase 4: Execution & Review (Eksekusi & Review)

Tulis kode secara presisi dan bedah, jalankan unit testing 100% lolos, dan lakukan review keamanan.

Agent: /sdlc-write-code, /sdlc-code-review Dokumen Input: Approved Plan 
/sdlc-write-code implement the shopping cart based on @plan-shopping-cart.md, and ensure all tests pass
/sdlc-code-review review my service layer and suggest refactoring

Phase 5: Documentation & RCA (Dokumentasi & RCA)

Tulis dokumentasi terstruktur kerangka Diataxis dan lakukan RCA mendalam jika terjadi bug fungsional.

Agent: /sdlc-generate-docs, /sdlc-bug-report Dokumen Input: PRD / Spec / Error Log 
/sdlc-generate-docs write an API reference guide based on @spec-shopping-cart.md and @cart.js
/sdlc-bug-report analyze the bug report in @issue-123.md and propose a fix for @cart.js

Protokol Injeksi Konteks

Untuk mencegah hilangnya konteks, halusinasi, dan untuk menegakkan keterlacakan SDLC yang ketat, Anda WAJIB melampirkan secara eksplisit, menyebutkan (misalnya menggunakan @filename), atau menyediakan dokumen hulu yang diperlukan dalam konteks prompt saat memanggil agen. Anda juga sangat disarankan untuk menyertakan file atau potongan kode relevan lainnya untuk melengkapi analisis.

Agen / Fase Dokumen Pendukung Wajib
/sdlc-map-architecture Basis kode sumber / konfigurasi build
/sdlc-draft-prd Project Discovery Draft (atau PRD yang ada)
/sdlc-clarify-reqs PRD, Spec, ATAU Plan (tergantung target)
/sdlc-define-specs PRD yang telah disetujui
/sdlc-plan-tasks Spesifikasi Teknis yang disetujui
/sdlc-write-code Rencana Implementasi ATAU Rencana Perbaikan Bug
/sdlc-code-review Spesifikasi Teknis DAN Rencana Implementasi
/sdlc-audit-consistency PRD, Spec, DAN Plan
/sdlc-generate-docs PRD, Spesifikasi Teknis, Rencana Implementasi, ATAU File Source Code Relevan

Opsi Jalur Cepat (Fast-Track): Untuk perbaikan minor, refaktorisasi, dan tugas ad-hoc, pemeriksaan dokumen wajib dapat di-fast-track. Agen secara proaktif menawarkan opsi ini agar Anda dapat melanjutkan tanpa seremonial lengkap SDLC.

Bypass SDLC (Pekerjaan Minor)

Untuk perbaikan kecil yang bersifat darurat (misal: mengganti teks, memperbaiki padding CSS, atau memperbaiki typo), Anda dapat melewati birokrasi SDLC penuh menggunakan perintah khusus /code-janitor.

Opsi Rekomendasi

1. Gunakan /code-janitor

Meskipun di-bypass, Anda tetap disarankan untuk melampirkan file yang akan dimodifikasi di dalam prompt obrolan Anda untuk memberikan konteks yang tepat. 

/code-janitor Tolong perbaiki padding tombol checkout di @style.css agar konsisten.
Opsi Alternatif

2. Gunakan Tag [Bypass SDLC]

Meskipun di-bypass, Anda tetap disarankan untuk melampirkan file yang akan dimodifikasi di dalam prompt obrolan Anda untuk memberikan konteks yang tepat. 

[Bypass SDLC] /sdlc-write-code perbaiki typo pada tombol submit di @header.html

Alur Kerja Mandiri dengan Slash Commands

Eksekusi seluruh siklus SDLC langkah demi langkah secara efisien menggunakan perintah slash langsung di ruang kerja AI Anda.

Pre-Phase: Architecture Memetakan Arsitektur & Topologi Basis Kode 
/sdlc-map-architecture scan this repository and map architecture into docs/ARCHITECTURE.md
Phase 0: Discovery Mengeksplorasi Codebase & Brainstorming 
/sdlc-explore-ideas explore the codebase and write a discovery draft for the new shopping cart feature based on @business-brief.md
Phase 1: PRD & Clarification Membuat PRD & Menginterogasi Ambiguitas 
/sdlc-draft-prd create a PRD for the shopping cart feature based on @discovery-draft.md
/sdlc-clarify-reqs interrogate the new @prd-shopping-cart.md for missing edge cases
Phase 2: Spec & Consistency Audit Menulis Spesifikasi Teknis & Audit Ketertelusuran 
/sdlc-define-specs design a technical specification based on @prd-shopping-cart.md
/sdlc-audit-consistency verify that @spec-shopping-cart.md strictly follows @prd-shopping-cart.md
Phase 3: Planning Membuat Rencana Implementasi Bertahap 
/sdlc-plan-tasks create a step-by-step implementation plan based on @spec-shopping-cart.md
Phase 4: Execution & Review Menulis Kode Fitur & Code Review 
/sdlc-write-code implement the shopping cart based on @plan-shopping-cart.md. Target files are @cart.js and @style.css
/sdlc-code-review review my service layer and suggest refactoring
Phase 5: Docs & Bug Fixing Menulis Dokumentasi & Diagnosis Bug 
/sdlc-generate-docs write an API reference guide based on @spec-shopping-cart.md and @cart.js
/sdlc-bug-report analyze the bug report in @issue-123.md and propose a fix for @cart.js
Test-First Engineering Suite

Full TDD-Spec SDLC Workflow

A rigorous 7-phase lifecycle combining pure Test-Driven Development (TDD), Behavior-Driven Development (BDD), Pre-Agreed Test Seams, and Floor-Guard anti-cheat enforcement.

7-Phase Pure TDD
AI Guide Need real-time workflow navigation? Run /tdd-ask-help anytime to diagnose workspace artifacts.
Auto-Routing
0

Phase 0: Architecture Mapping & Idea Discovery

Scans repository topography and evaluates initial ideas against measurable testable metrics.

/tdd-map-architecture map codebase
/tdd-explore-ideas explore @brief.md with measurable hypotheses
1

Phase 1: BDD Requirements & Doubt-Driven Clarification

Authors Given-When-Then user stories and interrogates assumptions until reaching Readiness Score >= 80.

/tdd-prd draft PRD with Given-When-Then scenarios from @discovery.md
/tdd-clarify interrogate @prd-cart.md for ambiguities (Score >= 80)
2

Phase 2: Contract Specifications, Test Seams & Impact Analysis

Designs exact TypeScript/DDL interfaces, maps public test seams, generates test-case checklists, and audits blast radius.

/tdd-spec define spec with Pre-Agreed Test Seams from @prd-cart.md
/tdd-checklist generate test inventory matrix from @spec-cart.md
/tdd-analyze audit blast radius and mocking boundaries @spec-cart.md
3

Phase 3: Tracer-Bullet Implementation Planning

Breaks specifications into vertical slices (DB to UI) with explicit RED-GREEN-VERIFY task sequences.

/tdd-plan-tasks plan vertical slices with RED-GREEN-VERIFY task sequence
4

Phase 4: Strict Red-Green-Refactor Coding

Autonomous execution strictly writing failing tests first, minimum code to pass, and refactoring cleanly.

/tdd-write-code execute @plan-cart.md with strict Floor-Guards
5

Phase 5: Audit Kualitas 5-Axis & Remediasi Bug Prove-It

Audit kode terhadap 5 sumbu kesehatan dan diagnosa bug dengan wajib menulis tes regresi gagal terlebih dahulu.

/tdd-code-review review cart service against 5-Axis Health Model
/tdd-bug-report diagnose @error.log using Prove-It failing test first
6

Phase 6: DiƔtaxis Living Documentation & Sprint Retro

Generates living documentation grounded in passing tests and optimizes test suite speed and reliability.

/tdd-generate-docs generate Diataxis docs with passing test examples
/tdd-retro analyze test runtime speed, flakiness, and update memory

Protokol Injeksi Konteks (TDD-Spec)

Untuk mencegah hilangnya konteks, halusinasi, dan menegakkan keterlacakan SDLC yang ketat, Anda WAJIB melampirkan secara eksplisit, menyebutkan (misalnya menggunakan @filename), atau menyediakan dokumen hulu yang diperlukan dalam konteks prompt saat memanggil skill TDD. Anda juga sangat disarankan untuk menyertakan log tes, kontrak antarmuka, atau error stack trace untuk melengkapi analisis.

Skill / Perintah Dokumen Pendukung Wajib
/tdd-initManifest repositori / CONSTITUTION / CONSTRAINTS yang ada
/tdd-map-architectureBasis kode sumber / konfigurasi test runner
/tdd-explore-ideasBrief pengguna / Konteks basis kode
/tdd-prdDraf Penemuan (docs/discovery/) ATAU PRD yang ada
/tdd-clarifyPRD, Spec, ATAU Plan (tergantung target)
/tdd-analyzeSpesifikasi Teknis ATAU Rencana Implementasi
/tdd-specPRD yang disetujui, ATAU User Brief Komprehensif
/tdd-checklistSpesifikasi Teknis yang disetujui ATAU Rencana Implementasi
/tdd-plan-tasksSpesifikasi Teknis yang disetujui (ATAU Plan yang ada)
/tdd-write-codeRencana Implementasi ATAU Rencana Perbaikan Bug
/tdd-code-reviewSpesifikasi Teknis DAN Rencana Implementasi
/tdd-bug-reportLaporan Bug / Error log / Stack trace
/tdd-generate-docsSpec, Plan, DAN file tes yang lulus
/tdd-generate-fixturesSpesifikasi Teknis (/spec/) DAN CONTEXT.md
/tdd-configure-ciCONSTRAINTS.md DAN Konfigurasi repositori
/tdd-mutation-testTest suite target DAN file kode implementasi
/tdd-retroLog eksekusi sesi DAN metrik durasi tes
/tdd-refactor-legacyFile kode legacy tanpa tes
/tdd-pair-coachTiket Rencana Implementasi ATAU kebutuhan Spec
/tdd-ask-helpPertanyaan pengguna ATAU kondisi repo (auto-deteksi fase)

Opsi Jalur Cepat (Fast-Track): Untuk perbaikan minor, refaktorisasi, dan tugas ad-hoc, pemeriksaan dokumen wajib dapat di-fast-track tanpa seremonial penuh SDLC dengan tetap menjaga disiplin test-first.

PĆ³lya Heuristic Engine

4-Stage Mathematical Problem-Solving Lifecycle

Inspired by George PĆ³lya's 1945 classic 'How to Solve It' and Uncle Bob's Clean Architecture, guaranteeing deep understanding before writing code.

1945 Heuristics & SOLID
Master Entrypoint Run /polya-router to auto-triage your workspace and determine the exact next step.
View Full Showcase
1

Stage 1: Understand the Problem

Analyze unknowns, verify condition data, explore existing codebase topography, and author definitive specifications with ambiguity elimination.

/polya-explore @brief.md
/polya-spec @docs/discovery/checkout-discovery.md
2

Stage 2: Devise a Plan (The Pause Rule)

Conceive a step-by-step strategy, slice into vertical tracer bullets, and enforce the mandatory Pause Rule to ensure user approval before execution.

/polya-plan @docs/spec/checkout-spec.md
3

Stage 3: Carry Out the Plan

Implement code incrementally with Micro-TDD discipline, maintaining single-responsibility and dependency inversion at every commit.

/polya-code @docs/plan/checkout-plan.md
4

Stage 4: Look Back (Review & Refactor)

Examine the result obtained, verify clean architecture integrity, conduct surgical root-cause analysis on bugs, and map living topography.

/polya-review @service.ts against @docs/spec/checkout-spec.md
/polya-docs @docs/spec/checkout-spec.md
/polya-fix analyze @bug-report.md