How abhisheksingh1180/jirapilot works
Project type and purpose
• JiraPilot is an automation/integration tool (open-source utility) whose goal is to listen for Jira events and automatically generate code via an LLM (Ollama) and push it back to GitHub.
• It is not a full-stack app but rather a “glue” script (code_generator.py) orchestrated by a GitHub Actions workflow (.github/workflows/codegen.yml).
File structure
• Top level:
• .github/workflows/codegen.yml – defines the event triggers and steps of the GitHub Action
• code_generator.py – Python entry point that implements the core logic: parsing Jira payload, calling Ollama API, writing files, committing and pushing
• requirements.txt – Python dependencies
• README.md / LICENSE – documentation
• There are no separate frontend/backend directories; the “architecture” is an event-driven pipeline.
README for additional insights
• Technology stack: Python, Ollama LLM, GitHub Actions, Jira Webhooks, GitHub REST API.
Detailed element checklist
• Jira → Runner (“Webhook: issue created”)
• Runner → Script (“Run codegen.py with payload”)
• Script → Ollama (“API call: prompt → code”)
• Script → Local FS (“Write code to /<path>”)
• Script → GitHub (“Check/version branch, commit & push”)
• External service vs. internal process vs. storage