Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.orxhestra.com/llms.txt

Use this file to discover all available pages before exploring further.

The prompt catalog builds structured system prompts from a PromptContext: 
from orxhestra import PromptContext, build_system_prompt

prompt = build_system_prompt(PromptContext(
    agent_name="WriterAgent",
    goal="Write polished articles from research notes.",
    instructions="Use markdown. Structure with headings. Be thorough.",
    skills=[
        {"name": "AP Style", "content": "Follow AP style guide for all prose."}
    ],
    agents=[
        {"name": "ResearchAgent", "description": "Retrieves research on any topic."}
    ],
    workflow_instructions=(
        "1. Call ResearchAgent to gather background information.\n"
        "2. Outline the article structure.\n"
        "3. Write the full article.\n"
        "4. Review and refine."
    ),
))

agent = LlmAgent(name="WriterAgent", model=model, instructions=prompt)
Sections are only included when non-empty - no boilerplate padding.