AI spelling and grammar checker for documentation

🐦 Twitter   •   📢 Discord   •   🖇️ AgentOps

Spellcaster is an open-source CLI tool that uses AI to improve your codebase's documentation. It scans repositories for:

• Grammar errors
• Spelling mistakes
• Issues in code examples

Spellcaster helps identify potential problems to help you write clear, concise, and professional documentation.

https://github.com/user-attachments/assets/60dbfdb1-d5a6-4dfd-9707-37e60d60b24e

To run Spellcaster:

pip install spellcaster spellcaster -d /path/to/your/docs/directory/

🦉 Features

• Grammar and Spelling Checks: Spellcaster reviews your documentation and comments to ensure they are clear, correct, and professional.
• Comprehensive File Scanning: Spellcaster reads across multiple file formats (Markdown, plain text, etc.) to detect issues.

Installation

• Install package:

pip install spellcaster

• Set up environment variables:

Create a .env file in your project root and add the following variables:

AGENTOPSAPIKEY=youragentopsapi_key
   ANTHROPICAPIKEY=youranthropicapi_key
   OPENAIAPIKEY=youropenaiapi_key
   GROQAPIKEY=yourgroqapi_key

Spellcaster uses AgentOps for cost and latency tracking. You can sign up for an API key here.

Replace your*api_key with your actual API keys for each service.

Usage

To run Spellcaster:

• In your terminal, use the following command structure:

spellcaster [options]

Options: - -d, --directory: The directory to scan - -u, --url: The GitHub repository URL to clone and scan - -l, --llm_provider: The LLM provider to use (choices: claude, sonnet, 3.5, gpt4o, gpt4, gpt3.5; default: MODEL from config) - -p, --proper_nouns: A string of proper nouns to include in the prompt (default: " Llama3.1-70B\n Cerebras\n Cohere\n OpenAI\n AgentOps\n Spellcaster") - -f, --filetypes: File types to scan (default: from FILETYPES in config) - -m, --maxfiles: Maximum number of files to scan (default: MAXFILES from config)

• Examples:

Scan a local directory with custom LLM provider:

spellcaster --directory /path/to/your/docs --llm_provider gpt4

Scan a GitHub repository:

spellcaster --url https://github.com/username/repo

• Spellcaster will analyze the specified directory or repository and output any detected issues, along with suggestions for fixing them.

Configuration

You can customize Spellcaster's behavior by adjusting the config.py file:

• FILE_TYPES: A list of file extensions to scan. By default, it includes .mdx and .md files. You can add or remove file extensions as needed.
• MAX_FILES: The maximum number of files to scan. By default, it's set to 500. You can change this number to suit your needs.
• MODEL: The default LLM provider to use. Model list available here.