================== /// MCP /// /// WRE /// ================== [server:online] [protocol:ready]
wren-engine
by Canner
The Semantic Engine for Model Context Protocol (MCP) clients and AI agents. Provides a semantic layer that lets LLM-powered agents query enterprise data sources with context, governance and security.
392
106
EnterpriseInstallation
1. Prerequisites
• JDK 17+ (wren-engine is written in Java)
• Maven 3.8+ or Gradle 7+ – either build tool works because the project ships with both wrappers.
• Docker (optional but recommended – an official container image is published on Docker Hub).
• A running database to register as a data source (PostgreSQL, BigQuery, Snowflake, etc.).
2. Clone the repository
git clone https://github.com/Canner/wren-engine.git
cd wren-engine
3. Build the server (choose one)
• Maven
./mvnw clean package -DskipTests
• Gradle
./gradlew clean build -x test
This produces `wren-engine-<version>.jar` inside `./target/` or `./build/libs/`.
4. Run locally
java -jar target/wren-engine-*.jar \
--spring.profiles.active=local \
--server.port=8080
The server will start listening on http://localhost:8080 .
5. (Optional) Run with Docker
docker run -p 8080:8080 \
-e OPENAI_API_KEY=YOUR_KEY \
-e DATABASE_URL=jdbc:postgresql://host:5432/db \
canner/wren-engine:latest
6. Initial configuration
• Copy `examples/application.yml` to `./config/application.yml` and enter:
- `openai.apiKey` (or other LLM provider key)
- Datasource connection strings
- Authentication secrets if you want to enable OAuth/JWT.
7. Verify
Open the swagger page at http://localhost:8080/swagger-ui or call the health probe:
curl http://localhost:8080/actuator/health
8. Deploy
The executable JAR or Docker image can be deployed to any JVM-compatible host, Kubernetes cluster or container platform. Helm charts are available under `deploy/helm` for K8s users.
Documentation
License: Apache License 2.0
Updated 7/30/2025