New: Statsig MCP Now Available in Scout
Experiment, analyze, and optimize with AI-powered feature management ⚡

We're thrilled to announce that Scout now supports Statsig MCP (Model Context Protocol) integration, bringing enterprise-grade feature management, experimentation, and analytics capabilities directly to your Scout agents!
What is the Statsig MCP?
Statsig is the modern feature management and experimentation platform trusted by companies like Notion, OpenAI, and Microsoft to ship features safely and measure their impact. With the new Statsig MCP integration in Scout, you can now create custom Scout agents that have direct access to your Statsig workspace through natural conversation.
Ready to get started? Join Scout today →
Complete Statsig MCP Tool Suite
The Statsig MCP integration provides your Scout agents with access to 40+ powerful tools that cover every aspect of feature management and experimentation:
🚀 Feature Gates & Rollouts
- Create Feature Gates: Build new feature flags with targeting rules, rollout percentages, and user segments
- Update Gates: Modify existing feature flags, change rollout percentages, and update targeting conditions
- Launch/Archive Gates: Safely launch features to production or archive deprecated flags
- List Gates: Browse and filter feature gates by status, rollout rate, tags, and creator
- Get Gate Details: Access comprehensive information about specific feature gates including rules and metrics
🧪 A/B Testing & Experiments
- Create Experiments: Set up sophisticated A/B tests with multiple variants, success metrics, and statistical controls
- Manage Experiment Lifecycle: Start, stop, abandon, or make decisions on running experiments
- Update Experiments: Modify experiment parameters, add metrics, and adjust allocation percentages
- Pulse Results: Retrieve real-time experiment results with statistical significance testing
- Experiment Groups: Enable/disable specific experiment groups for granular control
📊 Analytics & Metrics
- Create Custom Metrics: Define business metrics for measuring feature and experiment impact
- List Metrics: Browse all available metrics with filtering and search capabilities
- Metric Analysis: Access detailed metric definitions and historical performance data
- Event Tracking: Monitor user events and their relationship to features and experiments
👥 User & Team Management
- User Management: List users, invite team members, and manage permissions
- Team Operations: Create and manage teams with specific access controls and default settings
- Role Management: Define custom roles and permission structures
🏗️ Advanced Configuration
- Dynamic Configs: Create and manage dynamic configuration values that can be updated without code deployments
- Segments: Build user segments for precise targeting across features and experiments
- Layers: Organize experiments into layers for better conflict resolution
- Holdouts: Set up holdout groups to measure overall feature impact
Why This Integration Matters
Previously, managing feature flags and experiments required constant context switching between different tools and manual data analysis. Now, you can create specialized Scout agents that can:
- Serve as intelligent experimentation assistants that help design, launch, and analyze A/B tests
- Automate feature rollout decisions based on real-time performance data
- Generate comprehensive reports combining Statsig data with other integrated tools
- Act as data-driven product managers with full access to feature performance insights
- Provide real-time experiment monitoring and automated alerting
Start building your Statsig-powered agent →
Getting Started: Connect Statsig to Scout in 3 Easy Steps
Setting up Statsig MCP in Scout is incredibly straightforward:
1. Navigate to Integrations
In your Scout dashboard, click on "Integrations" in the main navigation menu.
2. Find Statsig MCP
Browse through the available integrations or use the search bar to find "Statsig MCP".
3. Connect the MCP
Once you've located Statsig MCP, simply click the "Connect MCP" button to begin the connection process.
That's it! Scout will guide you through the authentication process, and within minutes your Scout agents will have full access to your Statsig workspace data.
Create Custom Statsig-Powered Agents
Once connected, you can create specialized Scout agents with Statsig capabilities. Here are some powerful agent examples:
Experimentation Manager Agent
An agent that specializes in A/B testing and can:
- Design statistically sound experiments with proper sample sizes
- Monitor experiment health and statistical significance
- Generate experiment reports with actionable insights
- Recommend when to launch, iterate, or kill features
Feature Rollout Agent
An agent focused on safe feature deployment that can:
- Gradually increase feature rollout percentages based on metrics
- Monitor feature performance and automatically pause problematic rollouts
- Create targeted rollouts for specific user segments
- Generate rollout status reports for stakeholders
Product Analytics Agent
An agent that helps with data-driven product decisions and can:
- Track feature adoption and user engagement metrics
- Identify high-performing features and optimization opportunities
- Generate executive dashboards combining multiple data sources
- Provide recommendations for feature improvements
Join our Slack community for tips and best practices →
Chat Your Way to Better Feature Management
With Statsig MCP-enabled agents, you can have natural conversations about your features and experiments. Here are some examples of what you can ask:

Feature Management:
- "Show me all feature gates that are currently rolling out"
- "Create a new feature flag for the new checkout flow with 5% rollout"
- "What's the current rollout percentage for the mobile redesign feature?"
- "Archive all feature flags that haven't been updated in 6 months"
Experimentation:
- "Set up an A/B test for the new onboarding flow with conversion rate as the primary metric"
- "What are the results of the homepage experiment we launched last week?"
- "Create an experiment comparing GPT-4 vs GPT-3.5 for our chatbot"
- "Show me all experiments that are ready to make a decision"
Analytics & Insights:
- "What's our daily active user count for the past month?"
- "Create a custom metric to track feature adoption rate"
- "Show me which features are driving the most user engagement"
- "Generate a report on experiment velocity for Q4"
Team Coordination:
- "List all experiments owned by the growth team"
- "Create a new team for the mobile engineering group"
- "What features are currently being tested by the product team?"
Real-World Agent Use Cases
Here are some powerful ways teams are creating Statsig-powered Scout agents:
Growth Optimization Agent: Create an agent that continuously monitors conversion funnels, automatically sets up experiments to test improvements, and provides growth recommendations based on statistical analysis.
Product Manager Assistant: Build an agent that tracks feature performance across the entire product lifecycle, from initial rollout through full launch, providing data-driven insights for product decisions.
Engineering Safety Agent: Design an agent that monitors feature rollouts for performance regressions, automatically pauses problematic features, and alerts engineering teams to issues.
Executive Dashboard Agent: Develop an agent that generates executive-level reports combining feature adoption, experiment results, and business metrics for leadership reviews.
Customer Success Agent: Create an agent that tracks how new features impact customer satisfaction and retention, providing insights for customer success teams.
Compliance & Governance Agent: Build an agent that ensures all experiments follow proper statistical practices and company guidelines, maintaining experiment quality standards.
Building Your Statsig Agent
To create a Statsig-powered Scout agent:
- Navigate to Agents: Go to the Agents section in your Scout dashboard
- Create New Agent: Click "Create Agent" and give it a descriptive name
- Configure Capabilities: The Statsig MCP tools will be automatically available to your agent
- Customize Behavior: Set up your agent's personality, role, and specific Statsig-focused instructions
- Start a Session: You can now start a session with your agent and begin managing features through conversation
Ready to build your first Statsig agent? Get started here →
Advanced Statsig Capabilities
The integration supports Statsig's most advanced features:
Statistical Rigor
- Sequential Testing: Reduce experiment duration while maintaining statistical validity
- CUPED: Increase experiment sensitivity using pre-experiment data
- Multiple Comparison Corrections: Avoid false positives when testing multiple metrics
- Bayesian Analysis: Get probabilistic insights about experiment outcomes
Enterprise Features
- Warehouse Native: Connect directly to your data warehouse for custom metrics
- Team Management: Organize users into teams with specific permissions and defaults
- Review Workflows: Require approvals for sensitive feature changes
- Audit Logging: Track all changes for compliance and debugging
Advanced Targeting
- Custom Segments: Create sophisticated user targeting rules
- Holdout Groups: Measure overall feature impact across your product
- Layered Experiments: Run multiple experiments simultaneously without conflicts
- Dynamic Configs: Update configuration values without code deployments
What's Next?
The Statsig MCP integration is just the beginning. We're continuously expanding Scout's MCP ecosystem to help you build more powerful, specialized agents with access to all your favorite tools.
Ready to supercharge your feature management with Statsig-powered Scout agents? Head over to your Integrations page and connect Statsig MCP today, then create your first custom Statsig agent!
Have questions about the Statsig integration or ideas for other MCPs you'd like to see in Scout? Join our Slack community – we'd love to hear from you!