Looking for the self-hosted AI employee system? This page documents the open-source
pip install ragleap-terraform package. To install and run ragleap-core (AI employees, channels, Docker), see docs.ragleap.com; the project home is ragleap.com.Applied and destroyed on a real Docker Desktop host
ragleap-terraform
A Terraform module that creates a local kind cluster and installs the RagLeap Helm charts onto it. No cloud modules yet.
0.1.1
PyPI version
5 / 5
Chart pods Ready after apply
Install
The .tf files ship inside the wheel under ragleap_terraform/modules and ragleap_terraform/examples, and in the repo under packages/ragleap-terraform/.
pip install ragleap-terraform # or, with uv uv add ragleap-terraform
cd packages/ragleap-terraform/examples/kind terraform init terraform apply
Variables
cluster_nameDefault
ragleap-tf. Also namespace (default ragleap-core).install_observabilityDefault
true. observability_chart_path points at the chart directory.grafana_admin_passwordSensitive. Empty uses the chart's placeholder, which is not safe beyond a local test.
install_postgres_exporterDefault
false, see limitations.helm_timeoutSeconds, default 300.
kubeconfig_path keeps your own kubectl context untouched.Known limitations
- Local kind only. No AWS, GCP or Azure module exists.
- postgres-exporter is off by default. It needs the
ragleap-opsdatabase and the Secretragleap-db-exporter-secret, which no chart creates yet. This module does not installragleap-ops. - Credentials in plain text.
terraform.tfstateand the generated kubeconfig contain cluster credentials. Never commit them. - Never press Ctrl+C during apply or destroy. During testing it left a cluster running with no state. Clean up with
kind delete cluster --name ragleap-tf.