Context-Aware Agentic Troubleshooting System for the Resources AI Chatbot

[ajjaiahm]

What feature do you want to see added?

While exploring the Resources AI Chatbot plugin and running the project locally, I noticed that the current chatbot primarily focuses on retrieving Jenkins documentation and community resources using a RAG-based architecture.

Although this is very useful for documentation lookup, the chatbot does not yet utilize runtime Jenkins context such as, Build logs, Pipeline metadata, Plugin configuration, Job execution status

when i was analyzing the system architecture, I brainstormed a possible extension that could improve the chatbot’s effectiveness, accuracy, and operational usefulness.

The idea is to introduce a Context-Aware Agentic Architecture instead of a single retrieval pipeline, the chatbot could orchestrate multiple specialized agents responsible for different aspects of Jenkins knowledge.

**

**

Upstream changes

At this moment this idea primarily involves enhancements within the chatbot architecture itself.
future exploration may involve integrating Jenkins REST APIs to fetch runtime information such as build logs, pipeline metadata, and job execution status.

Are you interested in contributing this feature?

yess.. i im much interested in fixing problems and handle things at time with brainstorming and i love adding new innovative design thinking ideas so that it increase UX and move towards advanced mechnaisms

when studying the repository and running the plugin locally, I spent some time understanding how the chatbot backend works, especially the RAG-based workflow used for retrieving Jenkins documentation. This helped me understand how the current system retrieves knowledge and generates responses.
then I started thinking about how the chatbot could become more useful in real Jenkins workflows, especially for troubleshooting CI/CD pipelines.

The idea I brainstormed was to extend the chatbot so that it can also use runtime Jenkins context along with documentation retrieval. For example, integrating Jenkins APIs could allow the chatbot to access build logs, pipeline metadata, job status, or plugin configurations.

now i would like to experiment with:
• accessing runtime Jenkins data through Jenkins APIs
• designing a lightweight agent orchestration layer to route different types of queries
• exploring simple log analysis workflows for CI/CD troubleshooting
• improving context-aware retrieval so the responses become more accurate and practical