Your AI is breaking your rules
And you have no idea how often.
Dwellcount tracks when Claude, Cursor, and other AI tools violate your coding standards, ignore your CLAUDE.md, or break your .cursorrules. See every deviation. Fix the patterns.
Card required. Cancel anytime during trial.
You spent hours writing coding standards.
Your AI ignores them constantly.
CLAUDE.md
You wrote detailed instructions. Claude reads them. Then proceeds to use mongoose when you said native MongoDB driver only.
.cursorrules
“Always use TypeScript.” Cursor generates JavaScript anyway. “Prefer functional components.” Here's a class component.
Security Rules
“Never hardcode secrets.” AI generates apiKey = "sk-..." in the first suggestion. You catch it. Sometimes.
See every rule your AI breaks
Dwellcount watches your AI sessions in real-time. When it violates your standards, we catch it, categorize it, and show you the patterns.
Coding Standards
Style guides, naming conventions, file structure rules
Ex: "Use camelCase" → snake_case generated
Database Patterns
ORM usage, query patterns, connection handling
Ex: "Always use bulkWrite" → insertMany used
Security Rules
Secret handling, input validation, auth patterns
Ex: "Never hardcode keys" → API key in code
Architecture
Component structure, imports, dependency rules
Ex: "No circular imports" → circular detected
Your AI accountability dashboard
Finally see what your AI is actually doing to your codebase.
Recent Rule Deviations
5 minutes to full visibility
Add the MCP to your Claude config. That's it.
{
"mcpServers": {
"dwellcount": {
"command": "npx",
"args": ["@dwellcount/mcp"],
"env": {
"DWELLCOUNT_API_KEY": "your-api-key"
}
}
}
}Add the MCP
One line in your config. Works with Claude Desktop, Claude Code, Cursor, Cline.
Code normally
Use AI as you always do. Dwellcount watches silently, tracking every interaction.
See the truth
Dashboard shows every rule violation, cost, and productivity metric in real-time.
Beyond just catching violations
Full visibility into your AI-assisted development workflow.
Token & Cost Tracking
Know exactly how much your AI usage costs. Per session, per day, per project.
Rule Compliance Score
See your AI compliance rate over time. Track which rules get broken most.
Pattern Detection
Spot recurring violations. Fix the prompt, not the symptoms.
Tool Call Analytics
Which MCP tools get used? Success rates, duration, and failure patterns.
Team Insights
Compare AI usage across developers. Find who needs more guardrails.
Violation Alerts
Get notified when critical rules are broken. Catch security issues immediately.
Simple pricing. Serious value.
14-day trial. Per-seat pricing. One prevented incident pays for a year.
Pro
Full AI deviation tracking for professionals
$99/seat/mo
- Unlimited sessions
- Full rule deviation tracking
- Basic + Advanced AI models
- Cost & productivity analytics
- Pattern detection
- 90 day retention
- Email support
Enterprise
For teams with security requirements
$199/seat/mo
Min 3 seats
- Everything in Pro
- SSO / SAML
- API access
- Custom integrations
- 1 year retention
- Team comparisons
- Priority support
- Dedicated success manager
Built for developers who take AI seriously
If you're using Claude Code, Cursor, or any MCP-compatible tool daily, you need visibility into what it's doing.
Stop flying blind with AI
Your AI assistant is breaking your rules right now.
Find out how often.
Setup in 5 minutes. Cancel anytime.