R-A-C-E
A 4-dimension prompting framework to structure precise and actionable AI instructions: define a Role, specify the Action, provide Context, and frame the expected Execution.
Description
R-A-C-E structures your prompts across four dimensions: Role (the role the model must assume), Action (what it must do), Context (the information needed), and Execute (the expected output format and style). Its distinctive feature is the Execute dimension, which explicitly frames the output format (table, list, JSON...) for a directly usable result. This framework suits product teams, consultants, and entrepreneurs who use AI on a daily basis.
Objectives
- Improve team collaboration
- Structure AI prompts
How to apply R-A-C-E
- 1
R - Define the Role
Specify the profile and expertise level the model should adopt. The more specific the role, the more calibrated the response will be.
- 2
A - Specify the action (Action)
Describe what the model should do with a precise action verb: write, analyze, compare, summarize.
- 3
C - Provide the context (Context)
Give the necessary information: situation, constraints, target audience, available data, decisions already made.
- 4
E - Frame the Execution (Execute)
Define the output format: length, structure (list, table, prose...), tone, and language.