First PR Review

When you open a PR, Burg AI will automatically review it and post comments inline on the relevant lines of code.

What Triggers a Review

A pull request is opened
A pull request is updated (new commits pushed)
A draft PR is marked as ready for review

How Long It Takes

Review time depends on PR size, complexity, and current queue load. Times are approximate and may vary based on AI model response times and system load.

PR Size	Files Changed	Lines Changed	Typical Time
Tiny	1–3 files	< 100 lines	5–15 seconds
Small	4–9 files	100–500 lines	15–45 seconds
Medium	10–20 files	500–1,500 lines	45–90 seconds
Large	21–35 files	1,500–3,000 lines	90–180 seconds
Very Large	35+ files	3,000+ lines	3–10 minutes

Factors Affecting Review Time

File count and size: More files = more analysis time
Code complexity: Complex logic takes longer to analyze
AI model load: Peak hours may add 10–30 seconds
Queue position: Multiple PRs may cause brief delays
Content type: Binary files and very large files are skipped automatically

Pro tip: Reviews typically start within 5 seconds of PR creation, with most completing under 2 minutes during normal load.

Where Comments Appear

Burg AI posts review comments in the standard GitHub pull request interface. Comments appear in multiple locations for maximum visibility:

Comment Locations

Summary Comment — Overview of the entire PR review
- Appears in the Conversation tab as the main review comment.
- Provides PR overview, issue counts by severity, and key findings.
- Links to specific inline comments for detailed issues.
- Editable comment that gets updated if review is re-run.
Inline Code Comments — Posted directly on the specific lines of code with issues
- Appear in the "Files changed" tab.
- Hover over the comment icon to see the full review.
- Click to expand/collapse the comment thread.

Comment Structure

Burg AI uses a clear, structured format for all review comments to make feedback easy to understand and act on.

Severity Levels

Comments are categorized by severity to help you prioritize fixes:

🔴 Critical — Security vulnerabilities, data corruption risks, or system-breaking bugs that need immediate attention.
🟠 Major — Performance bottlenecks, logic errors, or maintainability issues that should be addressed.
🟡 Minor — Code style improvements, documentation gaps, or other non-blocking suggestions.

Content Elements

Issue Title — Clear, actionable description of the problem.
Technical Details — Why this is an issue and potential impact.
Code Context — Shows the problematic code with line numbers.
Fix Suggestion — Specific code changes recommended (when applicable).
Rationale — Explanation of why this matters for your project.

Interactive Features

Threaded Discussions — Reply to AI comments like any GitHub comment.
Resolve Comments — Mark issues as addressed when fixed.
Cross-references — Comments link to related issues in the same PR.
No Duplicate Comments — Same issues aren't flagged multiple times.

Example Review Comment

🔴 Critical: SQL Injection Risk

This query concatenates user input directly into SQL.
Use parameterized queries instead.

Suggestion:
- const query = `SELECT * FROM users WHERE id = '${userId}'`;
+ const query = 'SELECT * FROM users WHERE id = ?';
+ db.query(query, [userId]);

How to Re-Run a Review

Need to get a fresh review after making changes? Burg AI will automatically review new commits, but you can also trigger a re-review manually.

Ways to Trigger a New Review

Push new commits — The easiest way, just commit and push your fixes.
Close and reopen the PR — Forces a complete re-review of all files.

Skipped Reviews

A review may be skipped if:

You've hit your daily or monthly usage limit.
The PR exceeds file count or size limits.
All files are in excluded patterns (e.g., lock files, binaries, etc.).

Check the Troubleshooting guide if reviews aren't appearing.

Next Steps

Core Concepts — Understand how reviews work
Configuration — Customize review behavior

First PR Review

On this page