Prompt Expander Agent

Expand short prompts into detailed, actionable prompts for better task execution.

Turn a one-line request into a rich, actionable prompt with the PromptExpanderAgent.

from praisonaiagents import PromptExpanderAgent

agent = PromptExpanderAgent()

result = agent.expand("write a movie script in 3 lines")
print(result.expanded_prompt)

The Prompt Expander Agent transforms short, brief prompts into detailed, comprehensive prompts for better task execution. Unlike the Query Rewriter (which optimizes for search/retrieval), the Prompt Expander focuses on enriching prompts for task execution.

Agents: 1 — Specialized agent for prompt enhancement.

Quick Start

1. Simple Usage

   Expand a short prompt with the default strategy.

   from praisonaiagents import PromptExpanderAgent
   
   agent = PromptExpanderAgent()
   
   result = agent.expand("write a movie script in 3 lines")
   print(result.expanded_prompt)
   
2. With a Strategy

   Choose how aggressively to expand.

   from praisonaiagents import PromptExpanderAgent, ExpandStrategy
   
   agent = PromptExpanderAgent()
   
   result = agent.expand("AI blog", strategy=ExpandStrategy.STRUCTURED)
   print(result.expanded_prompt)
   

How It Works

Workflow

1. Receive short prompt
2. Optionally gather context via tools
3. Apply expansion strategy
4. Return detailed, actionable prompt

Setup

pip install praisonaiagents praisonai
export OPENAI_API_KEY="your-key"

Run — Python

from praisonaiagents import PromptExpanderAgent

agent = PromptExpanderAgent()
result = agent.expand("write a movie script in 3 lines")
print(result.expanded_prompt)

Run — CLI

# Expand a short prompt
praisonai "write a movie script in 3 lines" --expand-prompt

# With verbose output
praisonai "blog about AI" --expand-prompt -v

# With tools for context gathering
praisonai "latest AI trends" --expand-prompt --expand-tools tools.py

Run — agents.yaml

framework: praisonai
topic: Prompt Expansion
roles:
  expander:
    role: Prompt Expander
    goal: Transform short prompts into detailed prompts
    backstory: You are an expert at prompt engineering
    tasks:
      expand:
        description: Expand "write a movie script" into a detailed prompt
        expected_output: Detailed, actionable prompt

praisonai agents.yaml

Serve API

from praisonaiagents import PromptExpanderAgent

agent = PromptExpanderAgent()

# Note: PromptExpanderAgent uses .expand() method
# For API serving, integrate with standard agent

Expansion Strategies

BASIC

Simple expansion with clarity improvements. Fixes ambiguity and adds minimal context.

DETAILED

Rich expansion with context, constraints, format guidance, and quality expectations.

STRUCTURED

Expansion with clear sections: Task, Format, Requirements, Style, Constraints.

CREATIVE

Expansion with vivid, inspiring language and creative direction.

Basic Usage

from praisonaiagents import PromptExpanderAgent, ExpandStrategy

# Default (AUTO strategy)
agent = PromptExpanderAgent()
result = agent.expand("write a poem")
print(result.expanded_prompt)

Using Specific Strategies

from praisonaiagents import PromptExpanderAgent, ExpandStrategy

agent = PromptExpanderAgent()

# Basic - minimal expansion
result = agent.expand("AI blog", strategy=ExpandStrategy.BASIC)

# Detailed - rich context and requirements
result = agent.expand("AI blog", strategy=ExpandStrategy.DETAILED)

# Structured - clear sections
result = agent.expand("AI blog", strategy=ExpandStrategy.STRUCTURED)

# Creative - vivid language
result = agent.expand("AI blog", strategy=ExpandStrategy.CREATIVE)

Using Tools for Context

from praisonaiagents import PromptExpanderAgent

def search_tool(query: str) -> str:
    """Search for context."""
    # Your search implementation
    return "Latest AI trends: LLMs, multimodal, agents"

agent = PromptExpanderAgent(tools=[search_tool])
result = agent.expand("write about AI trends")
print(result.expanded_prompt)

Configuration Options

agent = PromptExpanderAgent(
    name="PromptExpander",
    model="gpt-4o-mini",
    
    temperature=0.7,      # Higher for creativity
    max_tokens=1000,
    tools=[...]           # Optional tools for context
)

ExpandResult Properties

result = agent.expand("write a poem")

# Access properties
print(result.original_prompt)   # Original input
print(result.expanded_prompt)   # Expanded output
print(result.strategy_used)     # Strategy that was used
print(result.metadata)          # Additional metadata

Convenience Methods

agent = PromptExpanderAgent()

# Direct strategy methods
result = agent.expand_basic("short prompt")
result = agent.expand_detailed("short prompt")
result = agent.expand_structured("short prompt")
result = agent.expand_creative("short prompt")

Key Difference from Query Rewriter

Feature	Query Rewriter	Prompt Expander
Purpose	Optimize for search/retrieval	Expand for task execution
Use Case	RAG applications	Task prompts
Output	Search-optimized queries	Detailed action prompts
CLI Flag	--query-rewrite	--expand-prompt

Example: Movie Script

Input:

write a movie script in 3 lines

Expanded (Creative Strategy):

Craft a captivating movie script distilled into just three powerful lines, 
each word infused with vivid imagery and emotional weight. Your lines should 
ignite the spark of adventure and intrigue, capturing a moment that hints at 
a grand journey ahead—one that resonates deeply with the audience's hearts 
and imaginations. Use poignant dialogue, evocative descriptions, and a 
tantalizing glimpse of conflict that will leave viewers breathless.

---

Monitor / Verify

praisonai "test prompt" --expand-prompt --verbose

Features Demonstrated

Feature	Implementation
Workflow	Single-step prompt expansion
Observability	--verbose flag
Tools	Optional context-gathering tools
Structured Output	ExpandResult with metadata

Best Practices

Match the strategy to the goal

Use BASIC to fix ambiguity, STRUCTURED for task specs, and CREATIVE for writing prompts. The wrong strategy over- or under-expands the request.

Raise temperature for creative expansion

Creative strategies benefit from a higher temperature. Keep it low for structured, deterministic expansions where you want repeatable output.

Attach tools to ground the expansion

Pass tools=[search_tool] so the agent pulls real context before expanding — useful when the prompt references current events or your own data.

Use Query Rewriter for search, not expansion

This agent enriches prompts for task execution. To optimise text for retrieval and RAG, use the Query Rewriter Agent instead.

Related

Optimise queries for search and retrieval.

Feed an expanded prompt into a research workflow.