Overview

The Editor is the space for managing your prompt engineering operations, which includes:

Selecting and configuring your preferred LLM.
Writing multi-modal, dynamic prompts using variables in the form of text, images, and PDFs.
Setting up MCP servers, tools, and response format for your prompts.

The Playground is the space for testing your prompts, which includes: * Running your prompts with your specific production test cases. * Going through all the runs’ history to iterate and restart from past runs, if they better fit the expected results. * Automatically running multi-turn chats by using auto tool calls, which allows the conversation to keep going.

The Variable Editor is the space that enables dynamic prompting, which includes:

Managing multi-modal variables (text, images, and PDFs) and setting their values.
Working with different prompt variable samples through linked datasets.
Linking other prompts to build agent systems with prompt chaining.
Fetching live data from external API endpoints at runtime.

Get started

Iterate

Evaluate

Deploy

Monitor

Guides

References

Key Features