)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":1,"name":"James E. Blair","email":"jim@acmegating.com","username":"corvus"},"change_message_id":"04a7ccb898ec4789296dae38c648880bf64b1869","unresolved":false,"context_lines":[{"line_number":23,"context_line":"with the workflow without explicitly making it *only* about"},{"line_number":24,"context_line":"LLM agents."},{"line_number":25,"context_line":""},{"line_number":26,"context_line":"For the record, I did not use an LLM to write this patch."},{"line_number":27,"context_line":""},{"line_number":28,"context_line":"Change-Id: I27ba603552434235dc634b46f53392da06b21794"},{"line_number":29,"context_line":"Signed-off-by: Monty Taylor \u003cmordred@inaugust.com\u003e"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"bb21b188_e0d0d373","line":26,"updated":"2026-04-22 22:17:37.000000000","message":"Thank you.  :)","commit_id":"7a222c406acc0eb93cf14f28b664fd419b4a3747"},{"author":{"_account_id":4146,"name":"Clark Boylan","email":"cboylan@sapwetik.org","username":"cboylan"},"change_message_id":"3ad212051662566d1551e513f69c29a89f2fae1b","unresolved":true,"context_lines":[{"line_number":15,"context_line":""},{"line_number":16,"context_line":"We can use Code-Review+1 to mark the initial assessment, which"},{"line_number":17,"context_line":"is a generally understood signal of \"someone has looked at this"},{"line_number":18,"context_line":"and found it ok\", so our overall semantics remain."},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"Add a pipeline that can be optionally used in place of the"},{"line_number":21,"context_line":"check pipeline that will wait for a Code-Review+1 vote in Gerrit."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"ced7fb3f_e3e48cae","line":18,"updated":"2026-04-23 15:28:04.000000000","message":"As we refine this I can imagine that we might care about who or what is leaving the Code-Review +1 (or eventual equivalent), but that specificity seems like an early optimization.","commit_id":"5d399a43c74f85ca67a1059d3489f990dc1a3c3a"},{"author":{"_account_id":2,"name":"Monty Taylor","email":"mordred@inaugust.com","username":"mordred"},"change_message_id":"8fe753eb06451745d5686a28da19225e2884fcc0","unresolved":true,"context_lines":[{"line_number":15,"context_line":""},{"line_number":16,"context_line":"We can use Code-Review+1 to mark the initial assessment, which"},{"line_number":17,"context_line":"is a generally understood signal of \"someone has looked at this"},{"line_number":18,"context_line":"and found it ok\", so our overall semantics remain."},{"line_number":19,"context_line":""},{"line_number":20,"context_line":"Add a pipeline that can be optionally used in place of the"},{"line_number":21,"context_line":"check pipeline that will wait for a Code-Review+1 vote in Gerrit."}],"source_content_type":"text/x-gerrit-commit-message","patch_set":3,"id":"8498442d_0121fc61","line":18,"in_reply_to":"ced7fb3f_e3e48cae","updated":"2026-04-23 15:45:11.000000000","message":"Agree. In my original thinking I was imagining an explicit label and tying that label to a gerrit group that the agents would be in. But as we\u0027ve talked things through, it feels like we can try this as a start and get more data, which is neat.","commit_id":"5d399a43c74f85ca67a1059d3489f990dc1a3c3a"}],"/PATCHSET_LEVEL":[{"author":{"_account_id":4146,"name":"Clark Boylan","email":"cboylan@sapwetik.org","username":"cboylan"},"change_message_id":"3ad212051662566d1551e513f69c29a89f2fae1b","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"4bfe0167_7080367f","updated":"2026-04-23 15:28:04.000000000","message":"FWIW when this was initially brought up I think I had in my head that Zuul would act as the mediator/broker for voting on Gerrit. I see now that the intent is more to have a separate system also listen for Gerrit events and take appropriate actions.\n\nI think this is fine and don\u0027t have concerns with that. However, it may still be advantages to using Zuul in this way as it gives our users a tool to configure and opt in/out of various actions using a system they are already comfortable configuring. That said there are probably advantages to not having Zuul sit between these systems as well. All things to think about as we learn more through experimentation. It will be interesting to hear about what you learn in the process.","commit_id":"5d399a43c74f85ca67a1059d3489f990dc1a3c3a"},{"author":{"_account_id":2,"name":"Monty Taylor","email":"mordred@inaugust.com","username":"mordred"},"change_message_id":"8fe753eb06451745d5686a28da19225e2884fcc0","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":3,"id":"70ce140e_8a0993be","in_reply_to":"4bfe0167_7080367f","updated":"2026-04-23 15:45:11.000000000","message":"Right now I have a long-running agent listening on stream-events separately. (actually, I have a server process that provides communication channels to agents listening. I agree, it could also be interesting to try modelling an amount of llm processing via zuul. Learning and experimentation FTW!","commit_id":"5d399a43c74f85ca67a1059d3489f990dc1a3c3a"}],"zuul.d/pipelines.yaml":[{"author":{"_account_id":1,"name":"James E. Blair","email":"jim@acmegating.com","username":"corvus"},"change_message_id":"04a7ccb898ec4789296dae38c648880bf64b1869","unresolved":false,"context_lines":[{"line_number":31,"context_line":"      Patchsets assesed by something (agent or human) enter this pipeline to receive an"},{"line_number":32,"context_line":"      initial +/-1 Verified vote."},{"line_number":33,"context_line":"    success-message: Build succeeded (vouched pipeline)."},{"line_number":34,"context_line":"    failure-message: Build failed (vouched pipeline)."},{"line_number":35,"context_line":"    manager: independent"},{"line_number":36,"context_line":"    precedence: low"},{"line_number":37,"context_line":"    require:"}],"source_content_type":"text/x-yaml","patch_set":1,"id":"345fed40_9b09f4cc","line":34,"updated":"2026-04-22 22:17:37.000000000","message":"Nit: These two lines are now the default!  You can omit them, but I don\u0027t think that\u0027s required for this change.  We can clean them all up later if we want.","commit_id":"7a222c406acc0eb93cf14f28b664fd419b4a3747"},{"author":{"_account_id":1,"name":"James E. Blair","email":"jim@acmegating.com","username":"corvus"},"change_message_id":"04a7ccb898ec4789296dae38c648880bf64b1869","unresolved":false,"context_lines":[{"line_number":46,"context_line":"          approval:"},{"line_number":47,"context_line":"            - Vouched: 1"},{"line_number":48,"context_line":"        - event: comment-added"},{"line_number":49,"context_line":"          comment: (?i)^(Patch Set [0-9]+:)?( [\\w\\\\+-]*)*(\\n\\n)?\\s*recheck"},{"line_number":50,"context_line":"    success:"},{"line_number":51,"context_line":"      gerrit:"},{"line_number":52,"context_line":"        # Note that gerrit keywords are case-sensitive."}],"source_content_type":"text/x-yaml","patch_set":1,"id":"02fd106f_68f43db9","line":49,"updated":"2026-04-22 22:17:37.000000000","message":"I agree it\u0027s fine to reuse \"recheck\" here, since we typically expect this pipeline to be mutually exclusive with check.","commit_id":"7a222c406acc0eb93cf14f28b664fd419b4a3747"}]}