Which LLM models are supported by these templates?

Kodexa prompt templates support major providers including Anthropic (Claude), OpenAI (GPT), and Amazon Bedrock, allowing for specific model IDs and temperature settings.

What is the benefit of using the knowledge system with prompts?

The knowledge system allows you to override generic prompts with specific logic or examples for particular vendors or document types without changing the underlying code.

Can I use variables within my prompt templates?

Yes, templates support Handlebars-style placeholders like {{content}} for document text, {{documentType}} for metadata, and {{taxonPath}} for semantic definitions.

What is a Kodexa prompt template?

A Kodexa prompt template is a YAML or Markdown definition that configures how an LLM should behave during document processing tasks like extraction, classification, or summarization.

How do I handle missing data in my extraction prompts?

The skill encourages defining explicit null-handling rules in the system prompt, instructing the model to return null for fields not found rather than inferring values.

Kodexa Prompt Template Authoring

Name: Kodexa Prompt Template Authoring
Author: kodexa-ai

bykodexa-ai

0•

Ciencia de Datos y ML

Authors and manages standardized LLM prompt configurations for document processing, data extraction, and AI-driven analysis on the Kodexa platform.

The Kodexa Prompt Template skill provides a structured framework for defining reusable LLM configurations, allowing developers to build sophisticated document processing pipelines with ease. It enables the creation of standalone YAML metadata or embedded markdown prompts within skill modules, supporting complex variable interpolation, multi-model support (Anthropic, OpenAI, Bedrock), and granular model tuning. By standardizing system and user prompts, developers can ensure consistent extraction, classification, and summarization outcomes while utilizing Kodexa's knowledge system for vendor-specific overrides and advanced prompt engineering best practices.

Características Principales

01Structured output configuration with JSON schema validation support

02Standardized YAML structure for system and user prompt definitions

030 GitHub stars

04Integrated knowledge system for per-vendor or per-document type overrides

05Multi-model provider support including Anthropic, OpenAI, and Bedrock

06Dynamic variable placeholders for document content, metadata, and taxonomy

Casos de Uso

01Building reusable system prompts for specialized AI assistants and skill modules

02Implementing multi-vendor processing logic using knowledge-based prompt overrides

03Creating high-precision extraction prompts for financial and legal document processing

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add kodexa-ai/kodexa-metadata-skills prompt-template

For use in Claude.ai and ChatGPT

Características Principales

01Structured output configuration with JSON schema validation support

02Standardized YAML structure for system and user prompt definitions

030 GitHub stars

04Integrated knowledge system for per-vendor or per-document type overrides

05Multi-model provider support including Anthropic, OpenAI, and Bedrock

06Dynamic variable placeholders for document content, metadata, and taxonomy

Casos de Uso

01Building reusable system prompts for specialized AI assistants and skill modules

02Implementing multi-vendor processing logic using knowledge-based prompt overrides

03Creating high-precision extraction prompts for financial and legal document processing