Package index • wizrd

Model Constructors Create agents for different LLM providers
`ollama_agent()` `llama_cpp_agent()` `openai_agent()` `azure_openai_agent()`	Agent Constructors
`llama()` `llamafile_llama()` `llama_vision()` `nomic()`	Convenience Functions for Common Models
`ollama_server()`	Construct an Ollama Server Object
Core Functions Main functions for working with LLMs
`chat` `predict.Agent` `predict.Chat` `chat.Agent` `chat.Chat`	Interact with an Agent
`last_output()` `last_message()`	Getting the Last Model Output
`LanguageModelParams()`	Language Model Parameters
Agent Grammar Functions to configure and customize agents
`instruct()`	Instructing Models
`prompt_as()` `system_prompt_as()`	Prompt Construction
`output_as()`	Output Constraints
`equip()` `unequip()`	Equipping Tools
`demonstrate()` `demonstrate_all()`	Demonstrating with Examples
Tools Tool calling and external integrations
`tool()`	Agent Tools
`describe_with_Rd()`	Describing Tools with Rd
`can_accept_as()`	Declaring Tool Parameters
Text Processing Text embedding and RAG functionality
`embed_text()`	Text Embedding
`text_store()`	Text Stores
`retrieve()`	Retrieve Documents
`chunk()` `default_chunking()`	Text Chunking
`rag_with()`	Retrieval Augmented Generation (RAG)
`persist()` `restore()`	Object Persistence
Model Context Protocol (MCP) Client Access tools, resources and prompts from an MCP server
`connect_mcp()` `start_mcp()` `tools()` `resources()` `prompts()`	Model Context Protocol (MCP)
Experimental Features that might change without notice
`c(<Agent>)` `c(<Chat>)` `c(<ChatPipeline>)`	Chat Pipelines

Reference

Model Constructors

Core Functions

Agent Grammar

Tools

Text Processing

Model Context Protocol (MCP) Client

Experimental