> ## Documentation Index
> Fetch the complete documentation index at: https://doc.entelligence.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Entelligence Code Review
> Catch production-impacting bugs with full-repository context, team guidance, actionable suggestions, and pull-request or local review workflows.
Catch production-impacting bugs before merge—not just style issues inside the diff. Entelligence reviews each pull request with the codebase, team guidance, prior feedback, and connected engineering context in view.
## Why full context changes the review
A function signature changes. The review notices the edit but misses callers in untouched files that still use the old arguments.
Entelligence traces the change across the repository, identifies affected callers, and points reviewers to the exact lines that can break.
The goal is a useful first review before a human spends time reconstructing the change from scratch.
## How a review works
Entelligence reads the pull-request diff and maps it to surrounding code, call sites, dependencies, and repository structure.
The review uses assigned guidelines, repository Markdown files, accepted Learnings, linked codebases, exclusions, and connected production evidence where available.
Checks cover correctness, security, performance, tests, edge cases, readability, and repository-specific standards. Findings are prioritized so merge-blocking risk does not get buried under minor feedback.
Entelligence posts the configured summary, inline findings, suggestions, and confidence signal on the pull request—or prints a local review in the terminal.
## What shows up on a pull request
| Output | What it gives the reviewer |
| ----------------------- | -------------------------------------------------------------------------------------- |
| High-level summary | A plain-English explanation of the change and the areas that deserve attention |
| Walkthrough | An ordered path through the implementation |
| Inline findings | Evidence beside the affected lines, with priority and reasoning |
| Committable suggestions | Ready-to-review code changes for findings with a clear fix |
| Prompt to fix with AI | Context you can hand to a coding agent without rebuilding the problem statement |
| Confidence score | A 1–5 readiness signal based on the review state |
| Changed-files summary | A file-level map of the change |
| Sequence diagrams | A visual explanation when the interaction is complex and diagram generation is enabled |
Entelligence accelerates the first review. Tests, required reviewers, branch protection, and your Git provider remain the merge authority.
## Tune signal without losing coverage
Choose critical-only, medium-and-above, or all findings per repository.
Add guidelines and Learnings so feedback reflects your architecture and conventions.
Choose inline or summary comments, explanation depth, generated content, and comment limits.
## Where Code Review runs
Automatically review ready-for-review pull requests on GitHub, GitLab, and Bitbucket, including supported self-hosted installations. Reviews can run when a pull request opens, updates, or is explicitly requested.
Run `entelligence review` before pushing. The same review engine can produce terminal output for developers, CI, or another coding agent.
Built-in language-aware guidance covers Python, JavaScript and TypeScript—including JSX and TSX—Java, Go, C, C++, C#, Ruby, Rust, and Dart. Cross-language checks apply beyond that list.
## Start reviewing
Connect source control, enable one repository, and verify the first review.
Learn how to triage feedback, apply suggestions, talk to the bot, and improve future reviews.