Overview
CodeAnt AI performs intelligent code reviews on every pull request, flagging issues across:- Syntactical errors
- Logical bugs
- Concurrency pitfalls
- Performance bottlenecks
- Security vulnerabilities
- Anti-patterns
Setup
-
Access Custom Prompt Settings:
- Navigate to
Settings→AI Code Review→Custom Prompts.
- Navigate to
-
Add a New Custom Prompt:
- Click “Add Custom Prompt” and define your custom logic in plain English.
- Example:
“Ensure that all new endpoints have proper authentication checks.”
-
Save and Apply:
- Click “Save” to activate the prompt.
- All future PRs will be evaluated against your custom logic in addition to default rules.
-
Explore Examples:
- Check out curated PR Custom Prompt Samples to get started quickly with prompt ideas.
Key Features
- Prompt-as-Policy: Codify organizational rules and domain knowledge directly into your code review pipeline.
- PR-Aware Evaluation: Prompts are applied only on the changed lines in the pull request to keep suggestions relevant.
- Natural Language Instructions: Write prompts in English—no scripting or DSL required.
- Comment-Level Feedback: Suggestions generated from custom prompts appear inline or as PR comments, just like default rules.
Benefits
- Flexible Review Framework: Adapt reviews to evolving team standards without modifying core CI pipelines.
- Faster Reviewer Onboarding: Prompts can embed tribal knowledge and legacy practices to help new reviewers follow along.
- Improved Consistency: Reinforces domain-specific expectations across all engineers and teams.
- Enhanced Compliance: Enforce coding contracts, licensing rules, naming conventions, or file structures unique to your business.