Overview

This an AI-powered tool that helps you generate Tekton PipelineRuns. It uses a combination of vector database storage and large language models to understand Tekton documentation and generate valid PipelineRun configurations and leverages AI agent to validate the generated PipelineRun YAML.

Features

• Document Ingestion: Automatically processes and indexes Tekton documentation for contextual understanding
• PipelineRun Generation: Creates Tekton PipelineRuns based on natural language requirements
• ChromaDB Backend: Uses ChromaDB vector database for efficient semantic search and retrieval

How It Works

The project uses Retrieval-Augmented Generation (RAG) to create robust Tekton YAMLs. When you run the generation script, the process follows these three steps:

• Query RAG: Loads the vector index from ChromaDB and queries it with your Jenkinsfile context.
• Retrieve Context: The system retrieves relevant code snippets and documentation chunks (based on similarity scores) from the ChromaDB vector database.
• Final Conversion: This retrieved context is sent to the Large Language Model (LLM) (Gemini) along with your conversion prompt, allowing the LLM to generate an accurate, context-aware Tekton PipelineRun YAML.

---

Prerequisites

• Podman (or Docker)
• A Gemini API key

Usage

1. Set up your API key

Export your Gemini API key as an environment variable:

export GEMINI_API_KEY="your-actual-api-key-here"

2. Start the Jenkins2Tekton container

The container comes with a pre-built RAG index, so you can start using it immediately:

podman run -it --rm \
    -e GEMINI_API_KEY="${GEMINI_API_KEY}" \
    -p 8320:8320 \
    quay.io/ngelman/jenkins2tekton:latest

Note: The pre-built index is included in the container image. You don't need to mount volumes or build the index yourself.

3. Generate Tekton Pipeline

Inside the container, copy your Jenkinsfile to /app and run the generation script:

# Copy your Jenkinsfile into the container (from another terminal):
podman cp /path/to/your/Jenkinsfile <container-id>:/app/

# Inside the container, generate the pipeline:
python generate_tekton_pipeline.py Jenkinsfile

Or mount your Jenkinsfile directory when starting the container:

podman run -it --rm \
    -e GEMINI_API_KEY="${GEMINI_API_KEY}" \
    -v $(pwd):/workspace:z \
    -p 8320:8320 \
    quay.io/ngelman/jenkins2tekton:latest

# Inside the container:
python generate_tekton_pipeline.py /workspace/Jenkinsfile

---

Development

For information on local development setup, testing, and contributing to the project, see DEVELOPMENT.md.