AI Prompt Collection

A system for saving, organizing, and retrieving AI prompts with metadata like tags, sources, and AI model compatibility information. Now includes a browser extension for capturing prompts while browsing and a local API service for seamless integration.

Features

• Save prompts with rich metadata
• Store source information (person, website, etc.)
• Tag prompts for easy categorization
• Track AI model compatibility
• Search by text or tags
• Command-line interface for management
• Browser extension for capturing prompts while browsing
• Local API service for programmatic access

Setup

Virtual Environment Setup

Following Python best practices, set up a virtual environment:

# Create a virtual environment
python3 -m venv .venv

# Activate the virtual environment
# On macOS/Linux:
source .venv/bin/activate
# On Windows:
# .venv\Scripts\activate

# Update pip
python3 -m pip install --upgrade pip

Dependencies

This project has minimal dependencies as it primarily uses Python's standard library.

JSON Structure

Prompts are stored in a JSON file with the following structure:

{
  "prompts": [
    {
      "id": "unique-id",
      "title": "Prompt Title",
      "content": "The actual prompt text...",
      "tags": ["tag1", "tag2", "tag3"],
      "source": {
        "type": "person/website/book/etc",
        "name": "Source name",
        "platform": "Where it was found",
        "date_added": "YYYY-MM-DD"
      },
      "ai_model": {
        "intended_for": ["model1", "model2"],
        "tested_on": ["model1"]
      },
      "metadata": {
        "created_at": "ISO timestamp",
        "updated_at": "ISO timestamp",
        "usage_count": 0,
        "effectiveness_rating": null,
        "notes": "Additional notes about the prompt"
      }
    }
  ]
}

Usage

Command Line Interface

The prompt_manager_v2.py script provides a command-line interface for managing your prompt collection.

Adding a Prompt

There are two ways to add prompts:

1. Using direct content input:

python3 prompt_manager_v2.py add \
  --title "My Prompt Title" \
  --content "The prompt text goes here..." \
  --source-type "person" \
  --source-name "John Doe" \
  --source-platform "Twitter" \
  --tags "tag1,tag2,tag3" \
  --models "chatgpt,claude" \
  --notes "This prompt is useful for X"

2. Using a file for content (recommended for longer prompts):

# First save your prompt to a text file
echo "Your prompt text here..." > prompt_content.txt

# Then add it to your collection
python3 prompt_manager_v2.py add \
  --title "My Prompt Title" \
  --content-file prompt_content.txt \
  --source-type "person" \
  --source-name "John Doe" \
  --source-platform "Twitter" \
  --tags "tag1,tag2,tag3" \
  --models "chatgpt,claude" \
  --notes "This prompt is useful for X"

Listing All Prompts

python3 prompt_manager_v2.py list

Getting a Specific Prompt

python3 prompt_manager_v2.py get <prompt-id>

Searching Prompts

# Search by text
python3 prompt_manager_v2.py search --query "learning"

# Search by tags
python3 prompt_manager_v2.py search --tags "education,interactive"

# Search by both text and tags
python3 prompt_manager_v2.py search --query "learning" --tags "interactive"

Programmatic Usage

You can also use the PromptManager class in your own Python scripts:

from prompt_manager_v2 import PromptManager

# Initialize the manager
manager = PromptManager()

# Add a prompt
prompt_id = manager.add_prompt(
    title="My Prompt",
    content="Prompt content...",
    source_type="website",
    source_name="example.com",
    tags=["tag1", "tag2"]
)

# List all prompts
prompts = manager.list_prompts()

# Get a specific prompt
prompt = manager.get_prompt(prompt_id)

# Search prompts
results = manager.search_prompts(query="learning", tags=["education"])

Components

1. Core Prompt Manager

The prompt_manager_v2.py script provides the core functionality for managing your prompt collection. See usage instructions below.

2. Local API Service

A FastAPI-based service that provides RESTful API endpoints for managing prompts. This allows other applications to interact with your prompt collection.

• Located in the api_service/ directory
• Runs locally on your machine (default: http://localhost:5000)
• Secured with API key authentication
• See API Service README for setup and usage

3. Browser Extension

A browser extension for Brave and Firefox that allows you to save prompts directly from web pages.

• Located in the browser_extension/ directory
• Select text on any webpage and save it as a prompt
• Add metadata like tags, source information, and notes
• Integrates with the local API service
• See Browser Extension README for installation and usage

Future Enhancements

• Web interface for easier management
• Integration with AI platforms for direct usage
• Prompt versioning and history
• Collaborative features for team usage
• Export/import functionality
• Advanced tagging and categorization
• Cloud synchronization options