Entelligence PR Reviewer

Entelligence PR Reviewer is an intelligent, automated code review assistant designed to streamline pull request (PR) evaluations across repositories. It enhances code quality, enforces consistency, and accelerates developer workflows with adaptive feedback rooted in your team’s unique coding standards and review practices.

Key Capabilities

• Automated Reviews: Provides insightful, production-grade comments on pull requests.
• Configurable Scoring System: Assigns impact, specificity, and urgency levels to prioritize critical feedback.
• Learning-Driven Adaptation: Adapts based on historical comments and explicit team feedback.
• Guideline Enforcement: Supports both autogenerated and uploaded team review guidelines.
• Review Enhancements: Auto-generated summaries, walkthroughs, and contextual replies.

PR Reviewer Settings

Comment Verbosity Control

Allows configuration of verbosity levels for review comments, ranging from brief summaries to comprehensive feedback.

Scoring Framework

Each pull request is evaluated on three dimensions:

1. Production Impact Score (PIS)

   Quantifies the potential runtime risk:

   • 5 — Definite production crashes/failures
   • 4 — High likelihood of issues
   • 3 — Moderate risk
   • 2 — Low risk
   • 1 — Minimal impact

2. PR Specificity Score (PSS)

   Scores how focused and applicable a PR is:

   • 5 — Clear, directly applicable fix
   • 3 — General refactor or improvement
   • 1 — Misaligned with scope

3. Urgency Impact Score (UIS)

   Prioritizes how soon the issue should be addressed:

   • 5 — Must fix immediately
   • 3 — Fix soon or in next sprint
   • 1 — Can be deferred

Recommendation: Start with a higher review threshold (score >14) and lower it as reviewer confidence increases.

Review Modes

Entelligence PR Reviewer offers two modes for comment verbosity:

• Concise – Provides minimalistic, to-the-point comments focused on clarity and brevity. Ideal for experienced teams or quick iterations.
• Verbose – Delivers detailed feedback with rich explanations, recommendations, and reasoning. Best suited for mentoring or complex code changes.

Review Features

• High-Level Summary: Generates an overview of PR changes.
• Walkthrough: Provides code navigation aid.
• Changed Files Summary: Lists impacted files.
• Auto-Generate Title: Suggests contextual PR titles.
• Auto-Link Ticket: Links to related issues.
• Auto Reply: Adds replies within threads.
• Enable LGTM Comments: Automatically approves clean PRs with LGTM remarks.

Learning Engine

Teach the Reviewer Your Preferences

Use inline commands to improve feedback accuracy:

@bot don't catch this

This instructs the system to suppress similar suggestions in future reviews. All feedback is stored and searchable in the "Learnings" tab.

Learning Configuration Options

The following learning features can be configured:

• Enable Sequence Diagram: Generate sequence diagrams for code changes.
• Cross Repositories Learnings: Enable learnings cross repositories.
• Cross Organization Learnings: Enable learnings cross organization.

Guidelines Management

Auto-Generate Guidelines

From the "Guidelines" tab, select a repository to analyze and synthesize historical comments into review standards.

Upload Your Own Guidelines

Upload .txt, .md, or .doc files (max 10MB) to enforce team-customized review protocols.

Configure Guidelines by Repository

Assign specific guidelines to codebases via the "Configure for Codebases" option.

Code Overview Dashboard

Provides at-a-glance analytics on pull requests:

• PRs Reviewed: Total number of evaluated PRs.
• Avg. Time to Merge: Time efficiency indicator.
• Acceptance Ratio: Percentage of PRs merged post-review.
• Review Comments: Total system-generated insights.
• Graphical Trends: Visualization of PR creation activity.
• PRs at Risk: Highlights overdue or stalled PRs.

PR Timeline

Chronologically displays merged, open, and closed PRs along with their author, comment count, and review sentiment (up/down votes).

Benefits

• Consistent Review Quality
• Automated Enforcement of Best Practices
• Reduced Manual Review Load
• Scalable Feedback Mechanism
• Data-Driven Improvement via Learnings and Metrics

Ideal Use Cases

• Large-scale teams requiring scalable review automation
• Projects needing adherence to security/performance standards
• Cross-repository consistency in feedback and code quality

Getting Started

1. Navigate to Pull Requests under "Code" on the left panel
2. Click Configure Repo
3. Install and authorize the Entelligence AI PR Reviews App to your GitHub repository
4. Create a pull request in your repository
5. The Entelligence AI Bot will automatically review your changes
6. Customize settings and preferences through the Entelligence Dashboard

Summary

Entelligence PR Reviewer transforms your code review process with AI-powered automation that learns from your team's practices. It provides intelligent, contextual feedback on pull requests while adapting to your coding standards over time. With configurable scoring systems, customizable guidelines, and comprehensive analytics, it helps teams maintain consistent code quality while reducing manual review overhead. The platform integrates seamlessly with GitHub and offers detailed insights into team performance and PR trends.