Dashboard Overview
Monitor your AI reply bot performance in real-time
Total Mentions
0
—
AI Replies Sent
0
—
Pending Review
0
Awaiting approval
Avg. Response Time
—
seconds / reply
Bot Status
MONITORING KEYWORDS
No keywords set — configure in Bot Setup
REPLY MODE
Manual Approval
AI PERSONA
Professional
Recent Mentions
No mentions yet
Connect your API keys and start the bot to monitor mentions.
Or click "Simulate Mention" to test.
Or click "Simulate Mention" to test.
Mention Monitor
Review, edit, and approve AI-generated replies
0 mentions
No mentions monitored yet
Start the bot from the Dashboard or use the Simulate button to test functionality.
Bot Setup
Configure keywords, persona, and behavior rules
Monitoring Rules
Define what tweets to monitor and reply to.
Safety & Limits
Manual Approval Mode
Review every reply before sending
Skip Retweets
Don't reply to retweets
Skip Replies
Only reply to original tweets
Duplicate Detection
Never reply to same user twice
AI Persona
Choose how your AI bot communicates with users.
QUICK PERSONA CHIPS
Professional
Friendly
Witty
Technical
Empathetic
Concise
Multilingual
Reply Templates
Add fallback templates when AI isn't sure.
Reply Analytics
Track performance and engagement patterns
Response Rate
0%
of mentions replied
Avg. Reply Length
0
characters
Auto-Approved
0
sent without review
Skipped / Rejected
0
not replied
📊 Replies by Day (Last 7 Days)
🎯 Sentiment Distribution
0%
POSITIVE
0%
NEUTRAL
0%
NEGATIVE
⚡ Top Mention Keywords
No data yet — start monitoring to see keyword frequency.
⏱️ Reply Time Distribution
No reply time data recorded yet.
Activity Log
Full audit trail of bot actions
No activity yet
Bot actions will appear here once you start monitoring.
API Settings
Manage your Twitter/X and Claude API credentials
Twitter / X API Keys
Get your API keys from developer.twitter.com. You need a Basic tier ($100/mo) or above for read/write access.
X API Basic: 500 tweets/month write limit. Upgrades available at developer.twitter.com
Claude AI (Anthropic)
Your Anthropic API key powers AI reply generation. Get yours from console.anthropic.com.
Claude Sonnet 4: ~$3 per 1M input tokens. Typical reply generation: <$0.001
Notification Settings
Browser Notifications
Alert on new mentions
Sound Alerts
Play sound on new mention
Log All Activity
Keep full audit trail
Data & Export
How to Use
Complete setup guide for Nova X AI Reply Bot
Quick Start (5 Minutes)
1
Get Twitter/X API Keys
Go to developer.twitter.com → Create App → Copy your API Key, API Secret, Access Token, and Access Token Secret. You need Basic tier ($100/mo) for write access.
2
Get Anthropic API Key
Visit console.anthropic.com → API Keys → Create new key. Copy the key (starts with sk-ant-...). Claude Sonnet 4 is recommended for best quality replies.
3
Enter Keys in Settings
Go to the Settings page → enter all 4 X API keys and your Anthropic key → Click "Save Settings". All keys are stored securely in localStorage.
4
Configure Bot Setup
Go to Bot Setup → add keywords to monitor (e.g., @yourhandle, #yourbrand) → choose persona style → set safety limits → Save Configuration.
5
Start the Bot
Go to Dashboard → click "Start Bot". The bot will begin polling X for new mentions every 60 seconds, generate AI replies via Claude, and queue them for your approval.
Important Limitations
X API Write Limit: Basic plan = 500 posts/month. Pro = 3,000/month.
Browser Limitation: This runs in your browser tab. Keep it open for continuous monitoring. For 24/7 operation, deploy the backend to a server.
Rate Limiting: X API v2 has strict rate limits. This bot respects them and adds delays automatically.
CORS Note: Direct X API calls from browser require a proxy. This app uses a relay for production use.
How the AI Reply Works
When a matching mention is detected, Nova X:
①
Fetches tweet content — reads the full tweet text, author info, and context thread
②
Sends to Claude API — with your brand context, persona, and the tweet as input
③
Generates contextual reply — under 280 characters, matching your persona and brand voice
④
Queues for approval — or sends automatically if auto-mode is enabled
⑤
Posts via X API — replies directly to the tweet with full tracking and logging
Best Practices
✅ Always start with Manual Approval Mode ON until you trust the AI output quality
✅ Review the first 20 replies manually to tune your custom prompt
✅ Set a blacklist of competitor names, offensive terms, and sensitive topics
✅ Keep max replies/hour at 10 or less to avoid X rate limits and spam flags
✅ Use specific keywords — broad terms = too many irrelevant mentions
⚠️ Never auto-reply to political or controversial topics
⚠️ X can suspend accounts for aggressive auto-reply behavior