Transform tweet agent into autonomous personality-driven continuous posting system

.env_example

@@ -55,6 +55,14 @@ TWITTER_USERNAME=your_twitter_username_here
 TWITTER_EMAIL=your_twitter_email_here
 TWITTER_PASSWORD=your_twitter_password_here
 
+# Twitter API v2 credentials (for automated posting with tweet_agent.py)
+# Get these from https://developer.twitter.com/en/portal/dashboard
+TWITTER_API_KEY=your_twitter_api_key_here
+TWITTER_API_SECRET=your_twitter_api_secret_here
+TWITTER_ACCESS_TOKEN=your_twitter_access_token_here
+TWITTER_ACCESS_TOKEN_SECRET=your_twitter_access_token_secret_here
+TWITTER_BEARER_TOKEN=your_twitter_bearer_token_here
+
 # google keys
 GOOGLE_APPLICATION_CREDENTIALS=your_google_credentials_path
 

docs/tweet_agent.md

@@ -1,37 +1,269 @@
-# Tweet Agent
+# Tweet Agent 🐦
 
-Generates trading-related tweets automatically.
+An autonomous AI agent that continuously generates and posts tweets in a customized personality-driven style.
 
 ## What It Does
-- Creates market commentary tweets
-- Generates token analysis threads
-- Writes educational content
-- Schedules tweet timing
+- 🤖 Generates tweets using AI (DeepSeek, Claude, or GPT)
+- 🔄 Runs continuously with configurable intervals
+- 🐦 Auto-posts to Twitter via API v2 (optional)
+- 💾 Saves all generated tweets to file
+- 🎯 Uses customizable personality manifest
+- ⚡ Handles errors gracefully and continues running
 
-## Usage
+## Features
+
+### Personality-Driven Writing
+The agent uses a detailed personality manifest to ensure consistent voice:
+- Casual, lowercase style (Moon Dev's voice)
+- Strategic emoji use (🚀 🌙 💰 🤖 ⚡ 🎯)
+- Tech jargon with simple explanations
+- Community-focused language
+- Focus on AI agents, trading, crypto, futurism
+
+### Autonomous Operation
+- Runs in continuous loop (default: 15 minutes per cycle)
+- Automatic tweet generation and posting
+- Keyboard interrupt support (Ctrl+C for graceful shutdown)
+- Error recovery without crashing
+
+### Safe Mode
+Works without Twitter API credentials - just saves tweets to file for review before manual posting.
+
+## Quick Start
+
+### 1. Basic Usage (No Twitter API - Safe Mode)
+```bash
+python src/agents/tweet_agent.py
+```
+
+This will:
+- Generate tweets every 15 minutes
+- Save to `src/data/tweets/generated_tweets_TIMESTAMP.txt`
+- NOT post to Twitter (requires API credentials)
+
+### 2. With Twitter Auto-Posting
+
+**Step 1: Get Twitter API Credentials**
+1. Go to https://developer.twitter.com/en/portal/dashboard
+2. Create a new app (or use existing)
+3. Generate API keys and access tokens
+4. Make sure you have "Read and Write" permissions
+
+**Step 2: Add Credentials to `.env`**
+```bash
+# Twitter API v2 credentials
+TWITTER_API_KEY=your_api_key_here
+TWITTER_API_SECRET=your_api_secret_here
+TWITTER_ACCESS_TOKEN=your_access_token_here
+TWITTER_ACCESS_TOKEN_SECRET=your_access_token_secret_here
+```
+
+**Step 3: Run the Agent**
 ```bash
 python src/agents/tweet_agent.py
 ```
 
-## Tweet Types
-- Market updates
-- Token discoveries
-- Technical analysis
-- Risk warnings
-- Educational threads
+The agent will now automatically post generated tweets to your Twitter account!
 
 ## Configuration
+
+All settings are in `src/config.py`:
+
+```python
+# Tweet Agent Settings
+TWEET_INTERVAL_SECONDS = 900  # 15 minutes between cycles
+TWEET_POST_DELAY_SECONDS = 10  # Delay between posting individual tweets
+
+# Twitter API Credentials (or use .env)
+TWITTER_API_KEY = ""
+TWITTER_API_SECRET = ""
+TWITTER_ACCESS_TOKEN = ""
+TWITTER_ACCESS_TOKEN_SECRET = ""
+```
+
+### Customizing the Interval
+
+To change how often tweets are generated, edit `src/config.py`:
+
+```python
+TWEET_INTERVAL_SECONDS = 1800  # 30 minutes
+# TWEET_INTERVAL_SECONDS = 3600  # 1 hour
+# TWEET_INTERVAL_SECONDS = 14400  # 4 hours
+```
+
+## Customizing the Personality
+
+The personality manifest is in `src/agents/tweet_agent.py` (lines 49-68):
+
+```python
+TWEET_PROMPT = """...
+
+Manifest:
+- casual, conversational tone with technical depth
+- lowercase writing style (no capitalization except for acronyms)
+- use emojis strategically (🚀 🌙 💰 🤖 ⚡ 🎯)
+- focus on ai agents, trading, crypto, and futurism
+...
+```
+
+Edit this manifest to change the agent's writing style!
+
+## Advanced Usage
+
+### Custom Text Input
+
+By default, the agent uses example text. To provide your own content:
+
+**Option 1: Direct text (edit the script)**
+```python
+test_text = """Your custom content here"""
+```
+
+**Option 2: Text file**
+1. Set `USE_TEXT_FILE = True` in `tweet_agent.py`
+2. Create file: `src/data/tweets/og_tweet_text.txt`
+3. Add your source content to that file
+
+The agent will chunk long text and generate tweets from each chunk.
+
+### Tweet Generation Settings
+
+In `src/agents/tweet_agent.py`:
+
 ```python
-TWEET_FREQUENCY = 'hourly'  # or 'daily', 'manual'
-TWEET_MAX_LENGTH = 280
-TWEET_INCLUDE_CHARTS = True
-TWEET_AUTO_POST = False  # Set True to auto-post
+MAX_CHUNK_SIZE = 10000  # Maximum characters per chunk
+TWEETS_PER_CHUNK = 3   # Number of tweets to generate per chunk
 ```
 
-## Safety
-- No financial advice disclaimers
-- No token shilling
-- Fact-based content only
+### AI Model Selection
+
+```python
+# In tweet_agent.py (line 14)
+MODEL_OVERRIDE = "deepseek-chat"  # or "0" to use config.AI_MODEL
+```
+
+Available models:
+- `"deepseek-chat"` - Fast & efficient (default)
+- `"deepseek-reasoner"` - DeepSeek's reasoning model
+- `"claude-3-haiku-20240307"` - Fast Claude
+- `"claude-3-sonnet-20240229"` - Balanced Claude
+- `"0"` - Use `config.AI_MODEL` setting
 
 ## Output
-`src/data/tweets/[date]/generated_tweets.txt`
+
+### File Output
+All tweets are saved to timestamped files:
+```
+src/data/tweets/generated_tweets_20250115_143022.txt
+```
+
+Format:
+```
+first tweet content here
+
+second tweet content here
+
+third tweet content here
+```
+
+### Console Output
+The agent provides detailed logging:
+```
+🔄 Starting new tweet generation cycle at 2025-01-15 14:30:22
+📊 Text Analysis:
+Total characters: 256
+...
+✅ Generated 3 tweets successfully
+📤 Posting tweet 1/3:
+✅ Tweet posted successfully! ID: 1234567890
+😴 Next run at: 2025-01-15 14:45:22
+```
+
+## Safety Features
+
+### Tweet Validation
+- ✅ Automatic length validation (280 char limit)
+- ✅ Truncation with ellipsis if too long
+- ✅ Character count displayed before posting
+
+### Error Handling
+- ✅ API errors don't crash the agent
+- ✅ Rate limit protection via delays
+- ✅ Network failures logged and retried
+- ✅ Graceful shutdown on Ctrl+C
+
+### Rate Limit Protection
+- Default 10-second delay between posts
+- Configurable via `TWEET_POST_DELAY_SECONDS`
+- Respects Twitter's API rate limits
+
+## Troubleshooting
+
+### "Twitter API credentials not found"
+- Add credentials to `.env` file
+- Agent will run in safe mode (file-only output)
+
+### "tweepy not installed"
+- Run: `pip install tweepy`
+- Or: `pip install -r requirements.txt`
+
+### "DeepSeek client not initialized"
+- Add `DEEPSEEK_KEY` to `.env`
+- Or set `MODEL_OVERRIDE = "0"` and use Claude/GPT
+
+### Tweets not posting
+1. Check Twitter API credentials are correct
+2. Verify app has "Read and Write" permissions
+3. Check console for error messages
+4. Ensure tweets are under 280 characters
+
+## Best Practices
+
+1. **Start in Safe Mode**: Run without Twitter API first to review generated tweets
+2. **Test with Short Intervals**: Use `TWEET_INTERVAL_SECONDS = 60` for testing
+3. **Monitor the Output**: Watch the first few cycles to ensure quality
+4. **Customize the Personality**: Edit the manifest to match your voice
+5. **Use Good Source Content**: Better input = better tweets
+
+## Example Workflow
+
+```bash
+# 1. Test tweet generation (no posting)
+python src/agents/tweet_agent.py
+# Review output in src/data/tweets/
+
+# 2. Add Twitter credentials to .env
+# 3. Test with short interval
+# Edit config.py: TWEET_INTERVAL_SECONDS = 120
+
+# 4. Run with auto-posting
+python src/agents/tweet_agent.py
+
+# 5. Monitor first few cycles
+# 6. Increase interval for production
+# Edit config.py: TWEET_INTERVAL_SECONDS = 3600
+
+# 7. Run long-term
+nohup python src/agents/tweet_agent.py &
+```
+
+## Integration with Other Agents
+
+The tweet agent can work with other agents in the ecosystem:
+
+- **Chat Agent**: Generate tweets from stream insights
+- **Research Agent**: Tweet about new trading strategies
+- **Sentiment Agent**: Tweet market sentiment analysis
+- **RBI Agent**: Share backtest results
+
+## Disclaimer
+
+⚠️ **Use Responsibly**
+- Review generated content before auto-posting
+- Ensure compliance with Twitter's Terms of Service
+- Don't spam or post misleading information
+- This is an experimental tool for educational purposes
+
+---
+
+*Built with love by Moon Dev 🌙 - Part of the Moon Dev AI Agents ecosystem*