PROMPT LIBRARY
⚡ 14 PROMPT
Kumpulan instruksi dan pola kerja LLM berdasarkan kebutuhan—mulai dari perencanaan spesifikasi, review kode, sampai aturan AGENTS.md.
Klik kartu untuk membaca konteks serta catatan penggunaannya, atau tekan tombol Salin untuk langsung menyalin prompt.
Status: “Pernah dipakai” berarti ada catatan hasil penggunaan nyata, bukan jaminan prompt akan selalu menghasilkan keluaran identik di semua model atau proyek.
Perencanaan
5 prompt-
Membuat rencana dari catatan diskusi
Mengubah catatan kasar hasil diskusi di TODO.md menjadi rencana implementasi yang menyeluruh lewat pertanyaan klarifikasi dari LLM.
Baca detail & catatan →Dear llm, take a look at TODO.md and ask a couple of questions. What should be included in this plan so that we can work towards something that we can implement and start from. -
Konfirmasi pemahaman dari ide kasar (stream of consciousness)
Memastikan LLM benar-benar menangkap maksud kita setelah menuangkan ide mentah secara panjang lebar (misalnya lewat voice-to-text) sebelum lanjut eksekusi.
Baca detail & catatan →[Tuangkan ide mentah atau transkrip voice-to-text di sini] --- restate to me what i just said in your own words so i know you understood me. -
Wawancara spesifikasi aplikasi satu pertanyaan per langkah
Meminta AI menggali ide aplikasi secara terstruktur lewat wawancara satu pertanyaan per satu waktu hingga siap menjadi spesifikasi teknis.
Baca detail & catatan →Ask me one question at a time so we can develop a thorough, step-by-step spec for this idea. Each question should build on my previous answers, and our end goal is to have a detailed specification I can hand off to a developer. Let’s do this iteratively and dig into every relevant detail. Remember, only one question at a time. Here’s the idea: [jelaskan ide aplikasi secara singkat] -
Kompilasi hasil diskusi menjadi dokumen spesifikasi lengkap
Mengubah hasil tanya-jawab brainstorming dengan LLM menjadi dokumen spesifikasi teknis lengkap (spec.md) yang siap diimplementasikan.
Baca detail & catatan →Now that we’ve wrapped up the brainstorming process, can you compile our findings into a comprehensive, developer-ready specification? Include all relevant requirements, architecture choices, data handling details, error handling strategies, and a testing plan so a developer can immediately begin implementation. -
Pecah spesifikasi menjadi blueprint dan rangkaian prompt bertahap
Memecah dokumen spesifikasi menjadi langkah-langkah iteratif kecil dan rangkaian prompt terintegrasi tanpa menyisakan orphan code.
Baca detail & catatan →Draft a detailed, step-by-step blueprint for building this project. Then, once you have a solid plan, break it down into small, iterative chunks that build on each other. Look at these chunks and then go another round to break them into smaller steps. Review the results and make sure that the steps are small enough to be implemented safely, but big enough to move the project forward. Iterate until you feel the steps are correctly sized for this project. From here, you should have the foundation to provide a series of prompts for a code-generation LLM to implement each step. Prioritize best practices and incremental progress, ensuring no big jumps in complexity at any stage. Make sure each prompt builds on the previous steps and ends with integration. No orphan code! Separate each prompt section using markdown. Each prompt should be tagged as text using code tags to ensure clear communication. < SPEC >
Review dan pemeriksaan
2 prompt-
Review dokumen secara menyeluruh
Meminta LLM memeriksa dokumen teknis atau spesifikasi secara menyeluruh untuk menemukan celah logika, kesalahan, dan bagian yang kurang tepat.
Baca detail & catatan →Please do check and review @phase-3-tools-and-web.md check for flaws, correctness and everything else. be thorough. -
Review kode untuk masalah konseptual dan kompleksitas
Memeriksa basis kode untuk menemukan dead code, kesalahan konseptual halus, inkonsistensi, abstraksi berlebihan, dan peluang kode idiomatis.
Baca detail & catatan →Please review all codes. Check for dead code, subtle conceptual errors, inconsistencies, overcomplicate code, bloat abstractions, inefficient and idiomatic [language] code
Implementasi
2 prompt-
Menyiapkan agent coordinator dan delegasi sub-agent di sesi Ampcode
Mengatur sesi utama sebagai koordinator yang mendelegasikan perencanaan ke planner thread dan mengeksekusi tugas paralel dengan sub-agent hemat di git worktree.
Baca detail & catatan →You are the long-term coordinator of the task. I believe these three tasks are doable by cheaper models, such as Haiku 5.5. Therefore, you should delegate each task to separate child in Claude Code mode with Haiku 5.5 model in the corresponding worktrees. However, we need a solid plan for them, right? So we should compose a decent plan, which may require another model. In this case, you may spawn another planner thread in high mode to compose the plan, then pass it to you for further coordination. -
Buat generator teks ke gambar
Contoh prompt satu berkas untuk membuat aplikasi web generator teks-ke-gambar menggunakan model Google Nano Banana lewat REST API tanpa SDK.
Baca detail & catatan →Create a text to image generator. The user will provide a text input. The Google Nano Banana Model will generate four different images. We will use the REST API. Do not use any SDK. Put everything in a single file. Also, have the ability for the user to provide the API key directly within the app. Make sure the app looks like it was designed by a billion-dollar design company.
Performa
1 promptInstruksi untuk asisten
2 prompt-
Rapikan README tanpa mengarang
Memperbaiki tampilan dan kelengkapan README repositori GitHub dengan tambahan badge dan cara memulai tanpa mengarang fitur.
Baca detail & catatan →You are a senior PM that is an expert with GitHub. You like good-looking repositories that have a lot of information on a project and how to get started. Spruce up this README and make it look nice with badges. Add missing content, don't embellish. Add emoji. -
Teknik Menghindari brainrot
Aturan AGENTS.md untuk mencegah utang kognitif dengan melarang agen mengedit berkas langsung sehingga developer mengetik setiap perubahan secara manual.
Baca detail & catatan →I want to understand every line of code that goes into this project. Never create, edit, move, rename, or delete project files unless I explicitly ask you to do so. Instead, show me every proposed edit in the chat so I can type it in manually. Do not run commands that modify project files, install dependencies, or change repository state unless I explicitly request that action. Instead, show me those commands in the chat so I can run them manually. I'm an experienced developer. Do not explain syntax, APIs, programming concepts, or implementation details unless explicitly asked.
Eksplorasi
1 promptPola kerja dan referensi
1 promptTidak ada prompt yang cocok dengan kata kunci tersebut.
Prompt dapat menghasilkan keluaran berbeda antar model, versi, dan konteks. Periksa hasilnya sebelum menggunakan atau membagikannya.
< KEMBALI KE BERANDA