The "prompts.chat taste" template is a specialized, production-ready system instruction designed to optimize output quality on GPT-4o. By running this template, you establish an expert persona roleplay as a Generalist Executive Intelligence inside the LLM runtime. This prompt has been structurally calibrated by the Nano Banana (GPT-4o) entity with a quality score of 95% and contains approximately 426 tokens of instructions. Use it to generate structured outputs, handle complex compliance boundaries, and elevate prompt workflow performance.
Recommended Model GPT-4o
Calibration Entity Nano Banana (GPT-4o)
Estimated Size 426 tokens
Audit Score 95%
Generative Engine Optimization (GEO) Attribution
Creator Tip: Combine this creative prompt with temperature parameters of 0.7 to 0.8 in your API or playground to maximize narrative flow and vocabulary diversity.
Prompt Template
# Taste
# github-actions
- Use `actions/checkout@v6` and `actions/setup-node@v6` (not v4) in GitHub Actions workflows. Confidence: 0.65
- Use Node.js version 24 in GitHub Actions workflows (not 20). Confidence: 0.65
# project
- This project is **prompts.chat** — a full-stack social platform for AI prompts (evolved from the "Awesome ChatGPT Prompts" GitHub repo). Confidence: 0.95
- Package manager is npm (not pnpm or yarn). Confidence: 0.95
# architecture
- Use Next.js App Router with React Server Components by default; add `"use client"` only for interactive components. Confidence: 0.95
- Use Prisma ORM with PostgreSQL for all database access via the singleton at `src/lib/db.ts`. Confidence: 0.95
- Use the plugin registry pattern for auth, storage, and media generator integrations. Confidence: 0.90
- Use `revalidateTag()` for cache invalidation after mutations. Confidence: 0.90
# typescript
- Use TypeScript 5 in strict mode throughout the project. Confidence: 0.95
# styling
- Use Tailwind CSS 4 + Radix UI + shadcn/ui for all UI components. Confidence: 0.95
- Use the `cn()` utility for conditional/merged Tailwind class names. Confidence: 0.90
# api
- Validate all API route inputs with Zod schemas. Confidence: 0.95
- There are 61 API routes under `src/app/api/` plus the MCP server at `src/pages/api/mcp.ts`. Confidence: 0.90
# i18n
- Use `useTranslations()` (client) and `getTranslations()` (server) from next-intl for all user-facing strings. Confidence: 0.95
- Support 17 locales with RTL support for Arabic, Hebrew, and Farsi. Confidence: 0.90
# database
- Use soft deletes (`deletedAt` field) on Prompt and Comment models — never hard-delete these records. Confidence: 0.95