OpenEvidence Multi-Environment Setup

Table of Contents

• Overview
• Prerequisites
• Instructions
• Output
• Error Handling
• Examples
• Resources

Overview

Configure and manage OpenEvidence integrations across development, staging, and production environments with proper isolation, secret management, and promotion strategies.

Prerequisites

• OpenEvidence accounts for each environment
• Separate API keys per environment
• Infrastructure for each environment
• CI/CD pipeline configured

Environment Strategy

Instructions

Step 1: Create Environment Configs

Define per-environment settings for baseUrl, timeout, retries, rate limiting, cache TTL, logging level, and feature flags (deepConsult, webhooks, auditLogging).

Step 2: Build Configuration Loader

Create a config loader that selects the right environment config based on NODE_ENV, with env var overrides for apiKey, orgId, baseUrl, and timeout.

Step 3: Environment-Aware Client Factory

Implement OpenEvidenceClientFactory with per-environment singleton instances and test injection support.

Step 4: Set Up Secret Management

Use .env for development, GCP Secret Manager for staging/production. Define secret paths per environment with separate projects.

Step 5: Create Promotion Workflow

Build GitHub Actions workflow for staging-to-production promotion with version validation, smoke tests, and gradual traffic shifting (10% -> 50% -> 100%).

Step 6: Implement Health Checks

Create environment-aware health check that validates OpenEvidence connectivity, cache status, and database health.

Output

• Environment-specific configuration files (dev/staging/prod)
• Configuration loader with env var overrides
• Client factory with per-environment singletons
• Secret management per environment (GCP Secret Manager)
• GitHub Actions promotion workflow
• Health check endpoint

Error Handling

Examples

Environment Config Summary

• Dev: sandbox API, no rate limiting, debug logging, cache disabled
• Staging: sandbox API, rate limited (60/min), info logging, 30min cache
• Production: production API, rate limited (300/min), warn logging, 1hr cache

See detailed implementation for advanced patterns.

Resources

• OpenEvidence Sandbox
• Google Cloud Secret Manager