proompt

A Go-based CLI tool that analyzes project structures to automatically generate concise project summaries as proompt.md files, optimized for copy-pasting into AI tools.

Problem Statement

Providing adequate context to AI tools about software projects presents several challenges:

Copying entire codebases exceeds context limits
Manual summarization requires significant developer time
Context boundaries limit comprehensive analysis
Project goals and structure are difficult to capture efficiently

Solution

proompt scans your codebase to:

Detect programming languages used
Analyze project structure
Identify key files and critical code snippets
Generate a concise proompt.md file with essential project context, formatted for seamless copy-pasting into AI platforms

Features

Language Detection: Automatically identifies programming languages and their distribution
Incremental Analysis: Uses Ollama queries to organize and summarize code
Context Optimization: Creates clear, concise documentation without unnecessary details, perfect for AI input
Key Snippet Extraction: Pinpoints critical code segments for relevant context
Project Goals Extraction: Identifies and summarizes project objectives
Root Integration: Places proompt.md at project root for easy reference

Implementation

The core implementation will be in main.go with these key components:

1. File System Traversal

Recursively walks project directories
Respects .gitignore patterns
Excludes binary files and dependencies

2. Language Detection

File extension analysis
Content pattern matching
Statistics on language distribution

3. Ollama Integration

Incremental queries to avoid context limits
Sequential refinement of project summary
Targeted extraction of project goals and key code snippets

4. Output Generation

Creates structured proompt.md
Includes language breakdown
Lists key project components and code snippets
Summarizes project goals
Formats content for direct copy-pasting into AI tools

Usage

proompt [project-path]

The tool will generate a proompt.md file in the root of the analyzed project with a concise summary and key code snippets, ready to be copied and pasted into any AI tool for enhanced context.

Next Steps

Implement file system traversal with proper .gitignore support
Develop language detection algorithms
Create Ollama integration for incremental analysis
Implement output formatting for proompt.md generation with AI-ready content

Why "proompt"?

Because sometimes, the best prompts come from your actual code, not your description of it. Let proompt prime your AI with precision context.

4cecoder / proompt