Aias00 opened a new issue, #959: URL: https://github.com/apache/rocketmq-dashboard/issues/959
## Problem `CliAgentProvider.complete` reads stdout and stderr sequentially before it calls `waitFor(timeout)`. A provider that fills stderr while stdout remains open can block both processes, so the configured timeout is never reached. `ClaudeCodeAgentProvider.stream` similarly consumes stdout without draining stderr. In addition, `OpenAiCompatibleLlmGateway` dispatches chat work through an unbounded cached thread pool. Slow or blocked CLI calls can create unbounded threads and degrade the Studio server. ## Scope Harden AI Native CLI execution for `claude-code` and `qoder`: drain child streams safely, enforce an end-to-end timeout, and use bounded execution with a structured saturation error. ## Acceptance criteria - Child stderr cannot block synchronous or streaming CLI execution. - The configured timeout covers the full child-process lifecycle. - Gateway task execution is bounded and rejects saturation predictably. - Add regression tests using a controllable test command or process abstraction. Track: RocketMQ Studio Track 3 / AI Native. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
