davsclaus opened a new pull request, #1853: URL: https://github.com/apache/camel-website/pull/1853
Part 2 of the chaos blog (part 1 is #1852): "let there be chaos", where an AI explores chaos on a running integration by itself for a time budget (`camel cmd chaos-explore`, CAMEL-26760), and reports what held and what broke. The post is built from a real recorded run: a local model (qwen3.6 35B via Ollama) explored the shop of the chaos example, with an added Kafka refunds route without error handling, for 10 minutes on 4.24.0-SNAPSHOT. - `let-there-be-chaos.gif`: a 16 second hook cut from the recording (the command, the refunds experiment failing, the AI's diagnosis, the saved report and scenario). - `chaos-exploration.svg`: the ten minutes as one figure, drawn from the saved exploration JSON: a time strip of the experiments, and per experiment what it did to each route. Columns 1 and 3 show the story: the same payment 503 fault, orders lose none, refunds lose 5 of 20. The red letters mark the verdicts that were misleading when the run was read afterwards. - A journal table of all ten experiments, the AI's diagnosis quoted, and an "Honest about the model" section. - `featured.jpg`: the part 1 image with a "Part 2" badge. Status: - `draft: true`, dated 2026-11-15, a week after part 1 (2026-11-08); move the date along if part 1 moves. - The features are in open PRs on apache/camel: #29498, #29499, #29501 and #29514 (stack), and #29512 (the dead-letter-channel count). Publish after they are merged. _Claude Code on behalf of Claus Ibsen_ 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_012nJsFsjqqFiwroeHRWaKbS -- 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]
