Documentation Index
Fetch the complete documentation index at: https://docs.praison.ai/llms.txt
Use this file to discover all available pages before exploring further.
Overview
PraisonAI v1.0 introduces a consistent Agent* prefix naming convention for core orchestration classes. This makes the API more intuitive and agent-centric.Backward Compatible: All old names continue to work as silent aliases. No code changes are required for existing projects.
What Changed
AgentTeam
Manages a team of agents working together on tasks.Old:
AgentManager, AgentsAgentFlow
Defines step-by-step workflows with routing and patterns.Old:
Workflow, PipelineAgentOSProtocol
Protocol for production deployment of agents.Old:
AgentAppProtocolAgentOSConfig
Configuration for AgentOS deployments.Old:
AgentAppConfigQuick Reference
| Old Name | New Name | Purpose |
|---|---|---|
AgentManager | AgentTeam | Multi-agent orchestration |
Agents | AgentTeam | Multi-agent orchestration |
Workflow | AgentFlow | Step-based workflows |
Pipeline | AgentFlow | Step-based workflows |
AgentAppProtocol | AgentOSProtocol | Deployment protocol |
AgentAppConfig | AgentOSConfig | Deployment config |
Migration Examples
- Python
- TypeScript
- AgentFlow
No Breaking Changes
All old names work exactly as before. They are silent aliases with no deprecation warnings.
Why the Change?
Consistency
Consistency
All orchestration classes now follow the
Agent* prefix pattern, making the API more predictable.Clarity
Clarity
- AgentTeam clearly indicates a team of agents
- AgentFlow clearly indicates a flow/workflow of agent steps
- AgentOS clearly indicates an operating system for agents
Agent-Centric
Agent-Centric
The naming reinforces that PraisonAI is an agent-first framework where everything revolves around agents.
Recommended Actions
Class Hierarchy
Need Help?
Documentation
Learn more about agents and orchestration
Examples
See working examples with the new names

