0.23.0 Updated UI and functionality

“𝗜𝘁 𝗼𝗻𝗹𝘆 𝗳𝗮𝗶𝗹𝘀 𝘄𝗶𝘁𝗵 𝘁𝗵𝗶𝘀 𝗼𝗻𝗲 𝗼𝗿𝗱𝗲𝗿.”

That’s the kind of Spring Boot bug I want an AI agent to help investigate.
The source code is available. The failing input, live bean state, and SQL behind that request often take more work to uncover.

Today I’m sharing fully updated 𝗦𝗽𝗿𝗶𝗻𝗴 𝗕𝗼𝗼𝘁 𝗥𝗘𝗣𝗟 𝗳𝗼𝗿 𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗝 — a work in progress that brings interactive exploration inspired by Lisp to running Spring Boot applications, with 𝗠𝗖𝗣 𝗮𝗰𝗰𝗲𝘀𝘀 𝗳𝗼𝗿 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀.

The workflow I’m building around:

𝗖𝗮𝗽𝘁𝘂𝗿𝗲 𝗮 𝗳𝗮𝗶𝗹𝗶𝗻𝗴 𝗰𝗮𝗹𝗹 → 𝗿𝗲𝗽𝗿𝗼𝗱𝘂𝗰𝗲 𝗶𝘁 → 𝗰𝗵𝗮𝗻𝗴𝗲 𝘁𝗵𝗲 𝗰𝗼𝗱𝗲 → 𝘃𝗲𝗿𝗶𝗳𝘆 𝘁𝗵𝗲 𝗳𝗶𝘅 → 𝗸𝗲𝗲𝗽 𝘁𝗵𝗲 𝘁𝗲𝘀𝘁.

📸 𝗦𝘁𝗮𝗿𝘁 𝘄𝗶𝘁𝗵 𝘁𝗵𝗲 𝗶𝗻𝗽𝘂𝘁 𝘁𝗵𝗮𝘁 𝗮𝗰𝘁𝘂𝗮𝗹𝗹𝘆 𝗳𝗮𝗶𝗹𝗲𝗱.

Capture method inputs and outcomes as DATA snapshots. Inspect the saved data and create variations to explore edge cases.

🔁 𝗚𝗶𝘃𝗲 𝘁𝗵𝗲 𝗮𝗴𝗲𝗻𝘁 𝗮 𝗿𝗲𝗽𝗲𝗮𝘁𝗮𝗯𝗹𝗲 𝗲𝘅𝗽𝗲𝗿𝗶𝗺𝗲𝗻𝘁.

Through MCP, Claude or Codex can inspect Spring beans, run permitted Java snippets, and use saved test cases (CASEs) to check expected behavior.

🔎 𝗙𝗼𝗹𝗹𝗼𝘄 𝘁𝗵𝗲 𝗲𝘃𝗶𝗱𝗲𝗻𝗰𝗲 𝗶𝗻𝘁𝗼 𝘁𝗵𝗲 𝗱𝗮𝘁𝗮𝗯𝗮𝘀𝗲.

Recorded calls, SQL activity, and Hibernate data help investigate unnecessary queries and potential N+1 problems.

🧪 𝗞𝗲𝗲𝗽 𝘁𝗵𝗲 𝗶𝗻𝘃𝗲𝘀𝘁𝗶𝗴𝗮𝘁𝗶𝗼𝗻 𝗮𝘀 𝗮 𝗿𝗲𝗴𝗿𝗲𝘀𝘀𝗶𝗼𝗻 𝘁𝗲𝘀𝘁.

Apply supported method-body changes with HotSwap, rerun the case, and export a JUnit test.
For an AI agent, the useful part is the feedback loop: inspect a real example, test a hypothesis, examine the result, and choose the next step.

𝗘𝘃𝗲𝗿𝘆 𝗲𝘅𝗽𝗲𝗿𝗶𝗺𝗲𝗻𝘁 𝗽𝗿𝗼𝘃𝗶𝗱𝗲𝘀 𝗻𝗲𝘄 𝗲𝘃𝗶𝗱𝗲𝗻𝗰𝗲 𝗳𝗼𝗿 𝘁𝗵𝗲 𝗻𝗲𝘅𝘁 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻.

📄 𝗜’𝘃𝗲 𝗮𝘁𝘁𝗮𝗰𝗵𝗲𝗱 𝘁𝗵𝗲 𝗘𝗻𝗴𝗹𝗶𝘀𝗵 𝗣𝗗𝗙 𝗴𝘂𝗶𝗱𝗲 with setup instructions, UI walkthroughs, and MCP usage.
I’d love feedback from Java developers