Boosting Log Parser plugins development for Log Parsing

Evaluation of Claude 3.7 in VS Code

1. Objective

The aim was to automate and accelerate the development of Log Parser Plugins by leveraging AI (Claude 3.7 | ZenCoder).

Manually creating similar parser plugins repeatedly was becoming repetitive and time-consuming.

Thus, the goal was:

• To automate the initial parser development using an AI assistant.
• To train AI with structured samples and clear instructions.
• To validate if the AI could produce production-ready Java plugins with minimal human correction.

AI Tools Used

• ZenCoder (Zencoder custom)

2. Approach

I followed a structured approach to enable Zencoder plugin to generate Log parser plugins:

2.1 Feeding Sample Data and Instructions

I prepared a folder structure containing the necessary resources for Claude to learn from:

• Prompt File:

  log_prompt.md

2.2 Prompt Given to Claude 3.7

I provided the following custom prompt to Claude: