Skip to content

CONFIG.md

Latest commit

 

History

History
93 lines (70 loc) · 3.53 KB

CONFIG.md

File metadata and controls

93 lines (70 loc) · 3.53 KB

Configuration Guide

This document explains all configuration options in config.json.

Model Configuration (model)

  • name: Primary LLM model for code review
  • fallback_models: Alternative models to try if primary fails
  • max_tokens: Maximum tokens for model response (increase for longer reviews)
  • temperature: Model temperature (0.0 = deterministic, 1.0 = creative)
  • timeout: API request timeout in seconds

Project Configuration (projects)

  • enabled_repos: List of repositories to enable AI review for (use ["*"] for all repos)
  • default_repo: Default repository for manual testing

Review Configuration (review)

  • max_diff_size: Maximum diff size to process in single review (characters)
  • large_diff_threshold: Threshold for large changes requiring chunked analysis
  • chunk_max_tokens: Max tokens per chunk in large diff analysis
  • max_files_detail: Maximum number of files to review in detail for large changes
  • overview_max_tokens: Max tokens for overview analysis of large changes
  • response_language: Response language code (see supported languages below)

Filter Configuration (filters)

  • ignored_extensions: File extensions to ignore during review
  • ignored_paths: Directory paths to ignore during review
  • code_extensions: File extensions considered as code files

Prompt Configuration (prompts)

  • include_line_numbers: Include line numbers in code analysis when possible
  • detailed_analysis: Provide detailed analysis for each section
  • security_focus: Focus on security vulnerabilities
  • performance_analysis: Include performance analysis

Supported Languages

Code Language Example
zh-TW Traditional Chinese 繁體中文
zh-CN Simplified Chinese 简体中文
en English English
ja Japanese 日本語
ko Korean 한국어
fr French Français
de German Deutsch
es Spanish Español
pt Portuguese Português
ru Russian Русский

Configuration Validation

The system automatically validates the configuration file structure and values on startup, including:

Structure Validation

  • Checks that all required configuration sections exist
  • Validates data types for each field
  • Ensures required fields are not empty

Value Validation

  • model.max_tokens > 0
  • model.temperature within 0.0-2.0 range
  • model.timeout > 0
  • All review section numeric parameters > 0
  • response_language must be a supported language code
  • enabled_repos format must be owner/repo or *
  • File extensions must start with .

Validation Error Examples

Configuration validation error: Missing required field: model.name
Configuration validation error: model.temperature must be between 0.0 and 2.0
Configuration validation error: Repository 'invalid-repo' must be in 'owner/repo' format or '*' for all repos
Configuration validation error: File extension 'txt' must start with '.'

Troubleshooting

JSON Syntax Errors

If the configuration file has JSON syntax errors, the program will report the error and exit on startup.

Configuration Validation Failures

The system will display specific validation error messages to help you fix configuration issues.

Model Connection Issues

Check that OPENAI_BASE_URL and OPENAI_KEY environment variables are correctly set.

Permission Issues

Ensure that GH_TOKEN has sufficient permissions to access target repositories.