How-to Guides#
Here you will find answers to “How do I…” questions. The proposed guides are goal-oriented and concrete, as they are meant to help you complete a specific task. Each code example in these how-to guides is self-contained and can be run with pyagentspec.
For conceptual explanations about Agent Spec and its components, see Agent Spec Specification.
For installation, see the Installation Guide.
For comprehensive descriptions of every class and function, see the API Reference.
Building Assistants#
Agent Spec provides a range of features to help you build two types of assistants: Agents and Flows.
These how-to guides demonstrate how to use the main Agent Spec features to create and customize your assistants.
- How to build a simple ReAct Agent
- How to connect MCP tools to assistants
- How to Develop a Flow with Conditional Branches
- How to Develop an Agent with Remote Tools
- Do Map and Reduce Operations in Flows
- How to Build an Orchestrator-Workers Agents Pattern
- Use OCI Generative AI Agents
- Use an A2A Agent
- Build a Swarm of Agents
- Build a Manager-Worker Multi-Agent System
- Build Flows with Structured LLM Generation
- Run Multiple Flows in Parallel
Additionally, we link the how-to guides offered by the WayFlow documentation. WayFlow is a reference runtime of Agent Spec, and among its how-to guides it proposes several examples of how to create common patterns using WayFlow and export them in Agent Spec.
Executing Assistants#
Agent Spec is framework-agnostic, and the assistants built using Agent Spec can be executed using any Agent Spec runtime. These how-to guides provide examples of how to run your Assistant using specific runtimes.
Configuration and State Management#
These guides demonstrate how to configure components in Agent Spec.
Ecosystem#
With expanding integrations and community partnerships, Agent Spec streamlines building, running, and monitoring agents across heterogeneous environments.
For more information about the Agent Spec ecosystem, see integrations and collaborations.
External Features#
pyagentspec enables the use of custom components in Agent Spec
configuration with the Plugin System. These how-to guides
provide examples of features that can be implemented with the plugin system.