implemented offline cache

README.md

@@ -68,8 +68,10 @@ daily-journal-prompt/
 ├── requirements.txt         # Python dependencies
 ├── ds_prompt.txt            # AI prompt template for generating journal prompts
 ├── historic_prompts.json    # History of previous 60 prompts (JSON format)
+├── pool_prompts.json        # Pool of available prompts for selection (JSON format)
 ├── example.env              # Example environment configuration
 ├── .env                     # Your actual environment configuration (gitignored)
+├── settings.cfg            # Configuration file for prompt settings and pool size
 └── .gitignore              # Git ignore rules
 ```
 
@@ -80,10 +82,12 @@ daily-journal-prompt/
 - **run.sh**: Convenience bash script for easy execution
 - **test_project.py**: Test suite to verify project setup
 - **requirements.txt**: Python dependencies (openai, python-dotenv, rich)
-- **ds_prompt.txt**: The core prompt template that instructs the AI to generate 6 new journal prompts
-- **historic_prompts.json**: JSON array containing the last 60 generated prompts
+- **ds_prompt.txt**: The core prompt template that instructs the AI to generate new journal prompts
+- **historic_prompts.json**: JSON array containing the last 60 generated prompts (cyclic buffer)
+- **pool_prompts.json**: JSON array containing the pool of available prompts for selection
 - **example.env**: Template for your environment configuration
 - **.env**: Your actual environment variables (not tracked in git for security)
+- **settings.cfg**: Configuration file for prompt settings (length, count) and pool size
 
 ## 🎯 Quick Start
 
@@ -140,14 +144,56 @@ python test_project.py
 
 ## 🔧 Usage
 
+### New Pool-Based System
+
+The system now uses a two-step process:
+
+1. **Fill the Prompt Pool**: Generate prompts using AI and add them to the pool
+2. **Draw from Pool**: Select prompts from the pool for journaling sessions
+
+### Command Line Options
+
+```bash
+# Default: Draw prompts from pool (no API call)
+python generate_prompts.py
+
+# Interactive mode with menu
+python generate_prompts.py --interactive
+
+# Fill the prompt pool using AI (makes API call)
+python generate_prompts.py --fill-pool
+
+# Show pool statistics
+python generate_prompts.py --pool-stats
+
+# Show history statistics
+python generate_prompts.py --stats
+
+# Help
+python generate_prompts.py --help
+```
+
+### Interactive Mode Options
+
+1. **Draw prompts from pool (no API call)**: Displays prompts from the pool file
+2. **Fill prompt pool using API**: Generates new prompts using AI and adds them to pool
+3. **View pool statistics**: Shows pool size, target size, and available sessions
+4. **View history statistics**: Shows historic prompt count and capacity
+5. **Exit**: Quit the program
 
 ### Prompt Generation Process
 
-1. The system reads the template from `ds_prompt.txt`
-2. It loads the previous 60 prompts from the fixed length cyclic buffer `historic_prompts.json`
-3. The AI generates 6 new prompts, attempting to minimize repetition
-4. Six new prompts are displayed.
-5. User selects one prompt for his/her journal writing session, which is added to the `historic_prompts.json` cyclic buffer.
+1. User chooses to fill the prompt pool.
+2. The system reads the template from `ds_prompt.txt`
+3. It loads the previous 60 prompts from the fixed length cyclic buffer `historic_prompts.json`
+4. The AI generates some number of new prompts, attempting to minimize repetition
+5. The new prompts are used to fill the prompt pool to the `settings.cfg` configured value.
+
+### Prompt Selection Process
+
+1. A `settings.cfg` configurable number of prompts are drawn from the prompt pool and displayed to the user.
+2. User selects one prompt for his/her journal writing session, which is added to the `historic_prompts.json` cyclic buffer.
+3. All prompts which were displayed are removed from the prompt pool permanently.
 
 ## 📝 Prompt Examples