[
https://issues.apache.org/jira/browse/CAMEL-25952?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Claus Ibsen resolved CAMEL-25952.
---------------------------------
Resolution: Fixed
Merged in https://github.com/apache/camel/pull/28269 (f4d7fae9f8bf)
_Claude Code on behalf of davsclaus_
> camel-jbang - AI tools: say where a language option such as resultType goes,
> and note a JMS reply header set by hand
> --------------------------------------------------------------------------------------------------------------------
>
> Key: CAMEL-25952
> URL: https://issues.apache.org/jira/browse/CAMEL-25952
> Project: Camel
> Issue Type: Improvement
> Components: camel-jbang
> Reporter: Claus Ibsen
> Assignee: Claus Ibsen
> Priority: Major
> Fix For: 4.24.0
>
>
> From the local-model benchmark of Oct 21-22 2026 (qwen3.6:35b-mlx through the
> camel-jbang-mcp server):
> *1. Where a language option goes.* A model wrote resultType next to the short
> form of the language:
> {code:yaml}
> - setBody:
> expression:
> jsonpath: "$[?(@.orderId == '${header.id}')]"
> resultType: java.util.Map
> {code}
> The validator answered only "property 'resultType' is not defined in the
> schema", and the model wrote the same file 23 times, into the next step. It
> now says where it goes: "resultType is an option of the jsonpath language:
> write the language as a map, with the text under expression: and resultType
> next to it (jsonpath: \{expression: "...", resultType: java.util.Map\})".
> This works for any language and any of its own options.
> *2. A JMS reply header set by hand.* The request-reply hint of CAMEL-25901
> never fired in the run: it is given for an unknown header that starts with
> Camel, and the models wrote JMSReplyTo, replyTo and
> $\{header.camelJmsReplyTo\}. A route that sets or reads a JMS reply header by
> hand, in a file where nothing is sent with InOut, now gets a note in the
> answers of the write and validate tools, and the write message points to it.
> It is a note, not an error: a reply queue set by hand, for a reply that
> another route consumes, is a design of its own.
> *3. Headers of an endpoint written as the scheme alone.* The header checks
> did not run for an endpoint written as uri: jms with its options under
> parameters:, the canonical form. They do now.
> _Filed by Claude Code on behalf of davsclaus_
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
