[Feature] Plan and execute module

[kapilreddy]

What feature would you like to see?

Implement a "Plan and Execute" module as part of DsPy's core library to support modern agent workflows that follow the plan-first, execute-later pattern.

ReAct

[LLM Call 1]
Thought: I need to find the current population of Tokyo
Action: search("Tokyo current population 2024")

[LLM Call 2] 
Observation: Tokyo metropolitan area has 37.4 million people
Thought: This seems like the metro area. Let me get the city proper population
Action: search("Tokyo city proper population 2024")

[LLM Call 3]
Observation: Tokyo city has 14.0 million people
Thought: Now I have both figures to give a complete answer
Action: respond_to_user

Plan and Execute

[LLM Call 1 - Planning Phase]
Thought: To answer about Tokyo's population, I should:
1. Search for current Tokyo population data
2. Clarify if they want city proper vs metropolitan area  
3. Get the most recent official figures
4. Present both if relevant

[LLM Call 2 - First Action]
Action: search("Tokyo population 2024 official statistics")

[LLM Call 3 - Second Action] 
Observation: Multiple results showing different boundaries...
Action: search("Tokyo metropolitan area vs city proper population")

[LLM Call 4 - Final Response]
Observation: Clear distinction found...
Final Response: [Complete answer with both figures]

Would you like to contribute?

• Yes, I'd like to help implement this.
• No, I just want to request it.

Additional Context

No response

[TomeHirata]

@okhat Do you have any insight on this feature request?