RepoTime-AI

• 

  uI

• 

  backend execution

• 

  backend post and get the requests

• 

Inspiration

CI pipelines fail repeatedly because developers unknowingly repeat mistakes hidden in repository history. RepoTime Machine uses AI to analyze past failures and predict future ones.

What it does

When a developer creates a merge request, the agent: • analyzes repository history • detects failure patterns • predicts CI risk • suggests improvements

How we built it

We created a flow of AI agents on the GitLab Duo Agent Platform: Repo Historian Agent Failure Pattern Agent Risk Predictor Agent Advisor Agent

Challenges we ran into

Real-time Risk Score Widget Displays current merge request risk score. Color-coded: Green (safe) → Yellow (moderate) → Red (high risk). Top Failure Patterns Graph showing recurring CI failure types. Hover shows the last commit that caused it. Energy & Sustainability Metrics (Green Agent) CI energy consumed per pipeline run. Recommendations for reducing energy.

Accomplishments that we're proud of

Predict CI failures before they happen.

• Automate actionable recommendations for developers.
• Visual dashboard for trends, patterns, and sustainability metrics.
• Green Agent for energy-efficient CI/CD pipelines.
• Easy integration with GitLab Merge Requests.

What we learned

Optional Modules

• Green Agent Module – Tracks CI/CD energy usage and recommends energy-efficient practices.
• Pipeline Optimizer – Suggests splitting large commits

What's next for RepoTime-AI

Built With

• ai
• charts
• cicd
• fastapi
• github
• gitlab
• javascript
• llm
• ml
• node.js
• postgresql
• python
• react
• sdlc
• typescript
• vscode