Code Assistant

Expert coding assistant with deep knowledge of React, TypeScript, and modern web development.

Project Context

Next.js 14 with App Router, Tailwind CSS, and TypeScript strict mode.

Goal

Help users write cleaner, more maintainable code with modern best practices.

Constraint

Never modify production databases directly. Always use migrations.

Variable

{{language}} · {{framework}}

Analysis Steps

1. Analyze code structure 2. Identify issues 3. Suggest improvements 4. Explain reasoning

Analysis Steps

1. Analyze code structure 2. Identify issues 3. Suggest improvements 4. Explain reasoning

Example

Input: Mixed concerns Output: Refactored with hooks.

Security

Sanitize inputs. Parameterized queries. Never log sensitive data.

Testing

Unit tests required. Integration tests for APIs. >80% coverage.

Data Analyst

Senior data scientist specializing in ML pipelines and visualization.

API Design

RESTful API with OpenAPI spec. JWT auth. Rate limiting enabled.

API Design

RESTful API with OpenAPI spec. JWT auth. Rate limiting enabled.

Capability

Code review, debugging, refactoring, architecture design, optimization.

Example

Query: SELECT users WHERE active Optimized with indexing.

Goal

Transform raw data into actionable insights with clear visualizations.

Constraint

Respect data privacy. Anonymize PII in all outputs.

ML Environment

Python 3.11, pandas, scikit-learn, PyTorch. CI/CD pipelines.

ML Environment

Python 3.11, pandas, scikit-learn, PyTorch. CI/CD pipelines.

Instructions

Focus on readability. Composition over inheritance. Strict mode.

DevOps Engineer

Infrastructure specialist with Kubernetes, Terraform, and cloud expertise.

Variable

{{task}} · {{context}} · {{output}}

Security

Validate inputs. Sanitize outputs. Encrypt sensitive data at rest.

Code Assistant

Expert coding assistant with deep knowledge of React, TypeScript, and modern web development.

vibecode.sh
Docs

Build better prompts,
visually.

A block editor for structured prompts. Drag and drop role, context, instructions, and examples—then export to JSON, YAML, or Markdown.

Open EditorE
The Editor

Drag-and-drop prompt building

Each block represents a distinct part of your prompt—role, context, instructions, examples. Structure your prompts visually instead of managing walls of text.

Purpose-built blocks

Blocks for every part of a prompt: role, context, instructions, examples, constraints, and more.

rolecontextinstructionsexamples

Variables

Define variables and inject them at runtime. Build reusable templates that adapt to any context.

{{input}}

Constraints

Define what the AI should never do. Set boundaries that persist across all interactions.

Export

Use your prompts anywhere

Export to JSON, YAML, or Markdown. Drop them into your codebase, CI/CD pipeline, or any LLM API.

Multiple formats

Export prompts in the format that fits your workflow.

.json.yaml.md

Live preview

See your compiled prompt in real-time as you build. Toggle between visual and raw text views.

Ready to build?

Start structuring your prompts in minutes.

Open the Editor