Stenographic Figure, Jackson Pollock (1943)
Museum of Modern Art (MoMA)
Abstract expressionism
![]() |
Juan Antonio Breña Moral Software Engineering Manager @ Capital One, C1
Twitter | Github | LinkedIn |
"Make it work, make it right, make it fast." - Kent Beck "Lead me, follow me, or get out of my way.", "Pressure makes diamonds." - George S. Patton Jr. |
jbang trust list
jbang cache clear
jbang catalog list jabrena
jbang qr-code@jabrena \
--url https://jabrena.github.io/cursor-rules-java/dvbe25/
Software Development workflow & Data Pipelines has evolved with the raise of AI tools enhancements.
For Software Engineers: AS IS
For Software Engineers: TO BE
For Data Pipelines: AS IS
For Data Pipelines: TO BE
For Data Pipelines: TO BE (II)
So... How to help Software Engineer & Data Pipelines to reduce ambiguity and improve the results of the different workflows?
Answer:
Use System prompts.
A prompt is an instruction, question, or input given to an AI system to generate a response. It's essentially how you communicate with AI to get it to do what you want.
Example:
Can you create a JMH benchmark
in order to know what is the best implementation?
A prompt could be enahanced with the help of a system prompt or a context.
A system prompt is a set of instructions given to an AI model that defines how it should behave. Think of it as the "operating manual" that shapes the AI's personality, capabilities, and boundaries.
A system prompt can include the following elements:
Role | Context | Goal |
Constraints | Instructions | Examples |
Output format | Safeguards | Metadata |
Valid for:
Anthropic Claude Sonnet,
OpenAI ChatGPT,
Google Gemini &
xAI Grok.
Examples:
128-java-generics.md
131-java-unit-testing.md
2003-agile-create-user-story.mdc
2006-adr-create-functional-requirements-for-rest-api-development.mdc
System prompts can help Engineering teams reduce ambiguity in the way that models answer their questions while increasing the homogeneity and consistency of results across different interactions and team members.
System prompts works in main Java IDEs/CLI tools:
![]() Cursor AI |
![]() Cursor CLI |
![]() Claude Code CLI |
|
![]() GitHub Copilot |
![]() GitHub Copilot CLI |
![]() JetBrains Junie |
But in Java IDEs, the AI tooling support for complex system prompts is not the same in all products.
Latest review: https://github.com/jabrena/cursor-rules-java/blob/main/documentation/reviews/review-20250829.md
For Data Pipelines, you could use CLI tools, and...
Cursor Background Agents REST API
https://cursor.com/docs/background-agent/api/endpoints
Your feedback is important to evolve this project. Participate in the following questionnaire:
Thanks