Skip to main content
Hierarchical workflows automatically handle common failure modes with built-in retry mechanisms and intelligent error recovery. 
from praisonaiagents import Agent, Task, AgentTeam

researcher = Agent(name="Researcher", instructions="Research topics thoroughly")
writer = Agent(name="Writer", instructions="Write from research")
team = AgentTeam(
    agents=[researcher, writer],
    tasks=[
        Task(description="Research AI trends", agent=researcher),
        Task(description="Write the article", agent=writer),
    ],
    process="hierarchical",
)
team.start("Write a blog post about AI trends")
The user starts a hierarchical team run; retries and manager re-prompts recover parse and assignment errors automatically.

​
Quick Start

1

Create Hierarchical Workflow

from praisonaiagents import Agent, Task, AgentTeam

# Create agents
researcher = Agent(
    name="Researcher",
    role="Research Analyst", 
    instructions="Conduct thorough research on topics"
)

writer = Agent(
    name="Writer",
    role="Content Writer",
    instructions="Write engaging content based on research"
)

# Create tasks
research_task = Task(
    description="Research AI trends",
    agent=researcher
)

write_task = Task(
    description="Write article about AI trends", 
    agent=writer
)

# Create team with hierarchical process (enables error recovery)
team = AgentTeam(
    agents=[researcher, writer],
    tasks=[research_task, write_task],
    process="hierarchical",
    manager_llm="gpt-4o"
)
2

Start Workflow

# Error recovery happens automatically
result = team.start()

# If manager has parse errors or selects invalid IDs,
# the system retries transparently
print(result)

​
How It Works

​
Configuration Options

SettingValueDescription
MAX_MANAGER_RETRIES3Maximum parse error retries
MAX_INVALID_SELECTIONS3Maximum invalid ID re-prompts
Backoff timing2s, 4s, 8sExponential backoff delays
These constants are not user-configurable today; they are SDK defaults designed for optimal reliability. Future versions may expose configuration options.

​
Error Recovery Types

Parse Failures

When: Manager LLM returns invalid JSON Recovery: Retry with exponential backoff Max: 3 attempts before abort

Invalid Task IDs

When: Manager selects non-existent task Recovery: Re-prompt with valid ID list Max: 3 attempts before abort

Loop File Errors

When: Cannot read loop input file Recovery: Mark task as failed with error Visibility: Error shown in TaskOutput

​
Best Practices

Isolate errors in parallel and loop steps so one failure does not abort unrelated work.
Return structured error details agents and UIs can display instead of silent exceptions.
Use before/after hooks to retry, notify, or escalate when steps fail.
Verify missing input files produce clear TaskOutput messages before production runs.

Hierarchical Workflows

Manager-based validation and graceful failure handling

Hooks

Custom error handling with hooks