Closes #3 and Closes #2

README.md

@@ -27,18 +27,20 @@ npm install -g sourcesailor
 
 #### Setup
 
-Set up the OpenAI API key and default model for SourceSailor:
+Set up the API keys and default models for SourceSailor:
 
 ```bash
-SourceSailor setup --apiKey <your_api_key> [--model <model_name>] [--analysisDir <directory>]
+SourceSailor setup --apiKey <your_api_key> [--model <model_name>] [--analysisDir <directory>] [--anthropicApiKey <your_anthropic_api_key>] [--geminiApiKey <your_gemini_api_key>]
 ```
 
 Options:
 - `--apiKey`, `-k`: OpenAI API Key (required)
-- `--model`, `-m`: OpenAI Model (default: 'gpt-3.5-turbo')
-- `--analysisDir`, `-a`: Root directory to write the analysis. Default is the home directory. Use 'p' to use the codebase directory. (default: home directory)
+- `--model`, `-m`: Default AI Model (default: 'gpt-3.5-turbo')
+- `--analysisDir`, `-a`: Root directory to write the analysis. Default is the home directory. Use 'p' to use the codebase directory.
+- `--anthropicApiKey`, `-n`: Anthropic API Key
+- `--geminiApiKey`, `-g`: Gemini API Key
 
-#### Get Directory Structure (Added in 1.3.0)
+#### Get Directory Structure
 
 Get the directory structure of the given path:
 
@@ -54,27 +56,26 @@ Options:
 - `--withContent`, `-c`: Include file content in the output (default: true)
 - `--ignore`, `-i`: Additional files or patterns to ignore for analysis. You can pass multiple patterns separated by commas (default: none)
 
-
 #### Analyze
 
 Analyze the given directory structure to understand the project structure and dependencies:
 
 ```bash
-SourceSailor analyse <path> [--verbose] [--openai] [--streaming] [--ignore]
+SourceSailor analyse <path> [--verbose] [--streaming] [--ignore] [--model]
 ```
 
 Positional arguments:
 - `<path>`, `-p`: Path to the directory to analyze (required)
 
 Options:
 - `--verbose`, `-v`: Run with verbose logging (default: false)
-- `--openai`, `-o`: Use OpenAI to infer project structure (default: true)
-- `--streaming`, `-s`: Use OpenAI streaming to infer project structure (default: false)
-- `--ignore`, `-i`: Additional files or patterns to ignore for analysis. You can pass multiple patterns separated by commas (default: none) (Added in 1.3.0)
+- `--streaming`, `-s`: Use AI streaming to infer project structure (default: false)
+- `--ignore`, `-i`: Additional files or patterns to ignore for analysis. You can pass multiple patterns separated by commas (default: none)
+- `--model`, `-m`: Specify the AI model to use for analysis
 
 #### List Models
 
-List all available OpenAI models:
+List all available AI models:
 
 ```bash
 SourceSailor listModels [--verbose]
@@ -96,16 +97,18 @@ Options:
 
 #### Update Config
 
-Update the OpenAI API key and default model:
+Update the API keys and default model:
 
 ```bash
-SourceSailor updateConfig [--apiKey <api_key>] [--model <model_name>] [--analysisDir <directory>]
+SourceSailor updateConfig [--apiKey <api_key>] [--model <model_name>] [--analysisDir <directory>] [--geminiApiKey <gemini_api_key>] [--anthropicApiKey <anthropic_api_key>]
 ```
 
 Options:
 - `--apiKey`, `-k`: OpenAI API Key
-- `--model`, `-m`: OpenAI Model
+- `--model`, `-m`: Default AI Model
 - `--analysisDir`, `-a`: Root directory to write the analysis. Default is the home directory. Use 'p' to use the codebase directory.
+- `--geminiApiKey`, `-g`: Gemini API Key
+- `--anthropicApiKey`, `-n`: Anthropic API Key
 
 #### Set User Expertise
 
@@ -122,7 +125,7 @@ This interactive command will guide you through setting your expertise levels fo
 Prepare a report based on the analysis:
 
 ```bash
-SourceSailor prepareReport <path> [--verbose] [--streaming]
+SourceSailor prepareReport <path> [--verbose] [--streaming] [--model]
 ```
 
 Positional arguments:
@@ -131,28 +134,30 @@ Positional arguments:
 Options:
 - `--verbose`, `-v`: Enable verbose output
 - `--streaming`, `-s`: Stream the output to a file
+- `--model`, `-m`: Specify the AI model to use for report generation
 
 Use the `SourceSailor --help` command to see the full list of available commands and options.
 
-
 ## About the Code
 
 The SourceSailor-CLI tool is structured around several key components, each serving a specific purpose in the code analysis process:
 
 - **Commands Directory**: Contains various CLI commands like `analyse.mjs`, `listConfig.mjs`, and more, which implement the tool's functionality.
-- **OpenAI Integration**: The `openai.mjs` module interfaces with the OpenAI API, enabling the tool to perform advanced code analysis.
-- ❌ :brain: **Tree-Sitter Parsing**: Utilizes Tree-Sitter grammars in `treeParser.mjs` and `treeSitterFromFieNames.mjs` for accurate code parsing across different languages. (Removed from the core functionality,See [#9](https://github.com/PrashamTrivedi/SourceSailor-CLI/issues/9))
+- **AI Integration**: Includes modules for OpenAI (`openai.mjs`), Google's Gemini (`gemini.mts`), and Anthropic (`anthropic.mts`) to perform advanced code analysis.
 - **Dynamic Command Handling**: Employs Yargs for building a flexible CLI interface, making it user-friendly and adaptable to various user needs.
 - **Configurable Analysis Directory**: Allows users to specify directories for storing analysis results, adding a layer of customization.
 - **User Expertise Levels**: The `expertise.mjs` module manages user expertise levels for different programming languages and frameworks, enhancing the tool's analysis capabilities.
+- **Model Utils**: The `modelUtils.mts` file provides a unified interface for managing different AI models and providers.
+
 
 ## :construction: Next Steps
 
 - Iterate through prompts (Will be ongoing evaluation)
 - Use Openrouter to switch models (Not planning to use, will switch to Gemini and Claude for future versions, may be by using plugins)
 - Use more prompts and CoT to work around the project (Will be ongoing evaluation) 
-- Add Anthropic and Gemini models (See [#3](https://github.com/PrashamTrivedi/SourceSailor-CLI/issues/3) and [#2](https://github.com/PrashamTrivedi/SourceSailor-CLI/issues/2) respectively)
-- ❌ Use tree sitter to get better ideas of the project and codebase :brain: (Can be core functionality of the CLI) (Already in the code) (Removed from the core functionality,See [#9](https://github.com/PrashamTrivedi/SourceSailor-CLI/issues/9) )
+- Add Plug and Play architecture for various model providers (Openrouter, AWS Bedrock, Google Vertex and Mistral are in my mind)
+- 🤔 Customizable prompts
+- ✅ Add Anthropic and Gemini models (See [#3](https://github.com/PrashamTrivedi/SourceSailor-CLI/issues/3) and [#2](https://github.com/PrashamTrivedi/SourceSailor-CLI/issues/2) respectively)
 - ✅ Write some reports per prompt and then pass it to CoT to generate a confident report (Already in the code)
 - Use the report as RAG. :bulb: (Highly speculative)
 - ✅ Use CLI decorators like colors and other decorators