Dependencies and kamelets
Weik.io 2026.2Add the libraries and kamelets an integration needs, and see where Weik.io resolves them from. Product page · Video
Watch: Connecting to another system, 4:30
The video calls another system through two declared Camel components and lists the project's kamelets.
This part of the video couldn't be loaded.
Weik.io resolves every jar of a version once, when it builds the version, so two agents on one version run the same jars. Because agents never download anything from Maven, an agent’s host needs no route to Maven Central and no build tools.
Camel components#
A component your routes use by URI, such as sql: or kafka:, needs no entry in integration.yaml. List a component there when the routes don’t show it, for example one that only a bean uses. A route whose address comes from a placeholder, such as to: "{{erp-url}}/api/invoices/pending", needs camel:http in the list, because the build can’t read the scheme of a placeholder:
name: erp-invoice-export
version: 1.0.0
dependencies:
- camel:jdbc
- camel:http
- mvn:com.microsoft.sqlserver:mssql-jdbc:12.8.1.jre11Other libraries#
Add a library from a Maven repository to dependencies as mvn:group:artifact:version, like mvn:com.microsoft.sqlserver:mssql-jdbc:12.8.1.jre11 above. You can also list libraries in camel.jbang.dependencies in application.properties.
Put a jar that is in no repository into the folder, for example in libs/. Every jar in the folder is on the classpath, whatever its size. When a new version keeps the same large jar, weikio apply doesn’t upload it again.
Maven repositories#
You don’t list Maven repositories in an integration. Weik.io fetches from the repositories set when it was installed and keeps their credentials as secrets. An installation that lists none fetches from Maven Central and from Atlassian’s public repository, which has the client that Camel’s Jira component needs.
An air-gapped installation fetches from a single mirror and never from Maven Central. Its weikio-backend init command sets the mirror with --maven-repository internal=https://maven.example.internal/repository/all,mirror and turns Maven Central off with --no-maven-central. Builds on both Camel lines work behind such a mirror with no internet access. Air-gapped installations has the setup.
Kamelets#
Weik.io looks for a kamelet in order and uses the first it finds:
- The integration’s folder: any
*.kamelet.yamlin the folder, which the integration uses without listing it. - The platform: the five kamelets every installation ships.
- The project: shared kamelets of the project, which an integration lists in
integration.yamlaskamelets: [erp-paging-source]. - Camel’s catalog: the kamelets of the catalog version pinned to the integration’s Camel line.
| Area | Platform kamelets |
|---|---|
| Channels | weikio-log, weikio-, weikio- |
| The entity store | entitystore-sink, entitystore- |
A project kamelet can’t have the name of a platform kamelet.
weikio get kamelets lists the project’s shared kamelets and the five platform kamelets:
$ weikio get kamelets --project financeSHADOWS marks a shared kamelet that has the name of a catalog kamelet. USED BY lists the versions that use each one.
In the admin UI, the Kamelets section of a project’s page has the same list and the YAML of each platform kamelet.
Kamelets for a local run#
weikio kamelets pull writes the project’s shared kamelets and the platform kamelets into a folder:
$ weikio kamelets pull ./kamelets --project financeThe four platform kamelets that work only on an agent get local stand-ins. weikio kamelets pull then prints the camel run line that uses the folder. Run that line in the integration’s folder when you can’t reach your Weik.io installation. When you can, weikio dev run does this for you.