Recent Revisions to This Document {#isv-doc-revisions}
======================================================

25.09.01
--------

:
Initial release.

LLM Quick Start Guide {#llm-quick-start-guide}
==============================================

This guide provides an introduction to getting started with Large Language Models (LLMs) using `Cybersource` documentation. Large Language Models have revolutionized how developers approach coding, documentation, and problem-solving. Using an LLM with `Cybersource` documentation enables you to find the information you need without finding and reading a specific guide.

Getting Started {#llm-quick-start-guide_getting-started}
--------------------------------------------------------

This section covers the fundamental concepts and initial setup required so that you can begin using LLMs effectively.  
When you start working with LLMs, you must:

* Understand model capabilities and limitations
* Select the appropriate model for your use case
* Set up proper authentication and API access
* Implement best practices for engineering prompts

Best Practices {#llm-quick-start-guide_best-practices}
------------------------------------------------------

To maximize the effectiveness of LLMs in your workflow, follow these established best practices:

1. Clear and specific prompts: Provide detailed context and specific instructions to get more accurate results.
2. Iterative refinement: Use follow-up prompts to refine and improve the initial output.
3. Validation and review: Always review and validate LLM-generated content before implementation.
4. Security considerations: Never include sensitive information such as payment information and merchant credientials in prompts or queries.
5. Version control: Track changes and maintain version history of LLM-assisted work.

Remember that LLMs are powerful tools that can augment human capabilities rather than replace human judgment and expertise.

Documentation Source {#llm-quick-start-guide_llm-documentation-source}
----------------------------------------------------------------------

Technical documentation for `Cybersource` is available in Markdown format. Markdown is easier for LLMs and AI agents to comprehend.  
All documentation appears here: [llms.txt](https://developer.cybersource.com/llms.txt "")  
To access a specific guide for use in your LLM, change the *.html* in a guide's URL to *.md* . For example, the Markdown version of [applepay-getting-started.html](https://developer.cybersource.com/docs/cybs/en-us/apple-pay/developer/ctv/rest/applepay/applepay-getting-started.md "") can be found at [applepay-getting-started.md](https://developer.cybersource.com/docs/cybs/en-us/apple-pay/developer/ctv/rest/applepay/applepay-getting-started.md "").

AI Agent Toolkit {#llm-quick-start-guide_agent-toolkit}
-------------------------------------------------------

`Cybersource` provides an agent toolkit at [agent-toolkit-intro.html](https://developer.visaacceptance.com/docs/vas/en-us/agent-toolkit/quick-start/all/na/agent-toolkit/agent-toolkit-intro.md ""). Use the agent toolkit to simplify the process of integrating `Cybersource` APIs in your agentic workflows.