Command reference

Creates a project, or adds Camel Kit to the current directory with --here. See Create a project for what is generated.

Configuration is read in three layers, the later overriding the earlier: the built-in distribution.properties, the user file given with -c or at its default path, and the -p flags. Any distribution property can be overridden; the common ones are camel.main.version, camel.springboot.version, springboot.bom.version, camel.quarkus.version, quarkus.platform.version, camel.mcp.version, knowledge.mcp.group-id, knowledge.mcp.version, citrus.version and citrus.mcp.version. The selected runtime and versions are persisted in .camel-kit/config.properties.

The default Knowledge runner is the published 0.0.2 artifact under io.github.luigidemasi. To use a migrated source build, override both knowledge.mcp.group-id=org.apache.camel and knowledge.mcp.version=0.0.3-SNAPSHOT. Changing the group alone would request an artifact that was never published.

init checks Java, JBang, Camel JBang and the Camel test plugin and reports missing tools as warnings; it never fails on them. On terminals with a native image protocol it shows a split-screen progress view, falling back to coloured text elsewhere.

Validates a generated workspace and prints PASS, WARN and FAIL findings with remediation text. Missing external prerequisites are warnings; broken generated files are failures and return exit code 1.

It checks .camel-kit/config.properties, the command stubs where the target uses them, the skill directories and manifests, the registered workspace templates, the exposure of user-invocable commands, the MCP filtering and permission fields the runtime supports, graph availability, the command prefix, the prerequisites and common stale references in generated files. Target specifics: for Codex it parses .codex/config.toml, the exact MCP allowlists, the prompt approval modes and the required fields of every custom-agent file; for Bob 2 it fails each public skill hidden by user-invocable: false; for Qwen it checks includeTools and trust; for OpenCode it evaluates all configuration layers as one and names the file that defines each rule. Configurations that predate the current filter and permission fields produce upgrade warnings; malformed current ones fail. Internal skills must exist as skills but must not be exposed as command stubs.

In the development snapshot, Citrus MCP configuration is optional. When the project has not opted in, doctor accepts a missing citrus entry; when opted in, a missing entry is a failure. Any existing Citrus entry is validated, including the tool filtering and permission fields supported by the target. An invalid entry produces a FAIL finding and exit code 1 even when the project has not opted in. These checks inspect configuration without starting the Citrus server or testing its availability. For compatibility, projects with citrus.mcp.version but no citrus.mcp.enabled are treated as opted in.

Creates the next numbered pipeline directory under docs/camel-kit/ and prints its ID. The slug is lowercase letters, digits and single hyphens.

Exit code 1 means an invalid slug or a directory that could not be created.

Manages the provenance and staleness of pipeline artifacts through YAML front matter.

Each artifact carries two blocks: generated (when, by which skill, from which upstream artifact; written once by doc init, which is idempotent) and staleness (written by doc stale and doc unstale). doc check prints the state as JSON and exits 0 whether or not the file is stale; a file without front matter reports stale: false. Every subcommand exits 1 on a missing file, an I/O failure or invalid arguments such as a blank --reason. doc stale always marks its target; with --cascade it also walks the sibling files whose generated.from names the target and marks them stale, recursively. After amending a design, mark each direct child separately (the implementation plan, and the migration runbook for migration pipelines) rather than the fresh design itself.

Builds and queries the project’s property graph, persisted at .camel-kit/project-graph.json.

See Graph intelligence for what each command returns.

Reads the yaml plan-metadata block of an implementation plan (falling back to the Markdown tasks for older plans) and prints the dependency waves and the task dependencies as JSON. /camel-execute runs it before dispatching tasks; see Waves.

Starts, inspects, resumes or aborts a Ship run (Technology Preview). The full option table, the oversight policies, the state and recovery rules and the examples are on the Ship page.