You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
wait for the asynchronously injected Team Leader context before test-18 reads the Leader AGENTS.md
when test-21 sees Leader activity only in the Leader DM, send one corrective nudge in the Team Room and mention the Leader there
keep the existing runtime-error and Team Room coordination assertions intact
Root causes observed in CI
test-18 could read Leader AGENTS.md before TeamReconciler's context write became visible in MinIO.
The Leader sometimes acknowledged the assignment in its DM but kept doing the work itself. A corrective DM message also stayed in that DM; placing the one-shot nudge in the Team Room gives the Leader the correct coordination context.
Temporary dependency
This branch is temporarily stacked on #985. The previous #1015 run's only two
failures were SHARD_A test-06-multi-worker: in both logs the Manager never
created Bob. #985 replaces that LLM-driven setup with deterministic controller
setup and is fully green across all 10 integration matrix jobs.
previous targeted SHARD_C (test-18) and SHARD_D (test-21) jobs passed
refreshed CI run 29250335473: all builds and all 10 integration matrix jobs passed, including both previously failing SHARD_A variants and SHARD_D/test-21
Fresh independent reproduction after the repository secret recovered: #1051 base-branch job 87604983224 reached test-21 and passed 43/44 assertions, but the Leader replied only in its DM (I need to include the roomId for the dev worker.) and never posted the Team Room assignment. The test then failed on the exact DM-only coordination path this PR handles with a one-shot Team Room nudge. This failure is unrelated to #1051’s GitHub token normalization and provides current-main evidence for #1015.
Test stabilization is useful, but the current stacked structure creates more maintenance cost than the test-only value justifies. #985, #1015, and #1007 share or depend on setup changes, and those commits are also leaking into unrelated PR diffs. Please flatten this into one independent test-only branch, or land the dependency first and then rebase each remaining PR so every diff contains only its own change. Also document the intended merge order if the PRs must stay separate.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
test-18reads the LeaderAGENTS.mdtest-21sees Leader activity only in the Leader DM, send one corrective nudge in the Team Room and mention the Leader thereRoot causes observed in CI
test-18could read LeaderAGENTS.mdbefore TeamReconciler's context write became visible in MinIO.Temporary dependency
This branch is temporarily stacked on #985. The previous #1015 run's only two
failures were SHARD_A
test-06-multi-worker: in both logs the Manager nevercreated Bob. #985 replaces that LLM-driven setup with deterministic controller
setup and is fully green across all 10 integration matrix jobs.
Relative to #985, this PR still changes only:
tests/test-18-team-config-verify.shtests/test-21-team-project-dag.shAfter #985 merges, this PR automatically reduces to those two files.
Validation
bash -n tests/test-06-multi-worker.sh tests/test-18-team-config-verify.sh tests/test-21-team-project-dag.shgit diff --checkagainst both currentmainand test: make multi-worker bob setup deterministic #985test-18) and SHARD_D (test-21) jobs passed