test(parallel-variable-key-lab): pin variableKey slots and the slot-aware parallel merge

[yilmaztayfun]

Summary

• New scenario parallel-variable-key-lab for fix(tasks): keep parallel task responses apart with variableKey and a slot-aware merge vnext#1072: the same SubProcess task (pvk-spawn-child, TaskType 14) runs twice at one order, each entry with its own variableKey, and a later entry with no variableKey keeps the legacy ToVariableName(task.key) slot.
• pvk-reuse pins the slot-aware merge: a parallel group at order 2 re-writes the slot order 1 left behind and overwrites it instead of faulting with Parallel tasks produced conflicting output.
• Three publish probes, built in-test so they never sit under core/, pin the publish rules: same order with no variableKey, the same variableKey twice, and a malformed variableKey are all rejected with 400.

Changes

• core/Tasks/parallel-variable-key-lab/: pvk-spawn-child (SubProcess → pvk-child), pvk-record-slots (Script).
• core/Workflows/parallel-variable-key-lab/: pvk-parent, pvk-reuse, pvk-child (all 1.0.0), generated by build-parallel-variable-key-lab.py from src/*.csx.
• tests/Core.IntegrationTests/Tests/ParallelVariableKeyLab/: ParallelVariableKeyLabTests (5 tests) + README.
• TEST-SCENARIOS.md: parallel-variable-key-lab row.

Test Plan

• npm run validate: all new components ✓ Valid. The total failure count (32 files) is unchanged from master.
• dotnet build tests/Core.IntegrationTests 0 errors.

Integration test evidence

• Scenario: Tests/ParallelVariableKeyLab (--filter FullyQualifiedName~ParallelVariableKeyLab).
• Runtime: burgan-tech/vnext feature/task-variable-key at 8046dbd6, then 8f11acfc after merging master; VNEXT_BASE_URL=http://localhost:4201 (run-docker.sh up core, local build).
• Result: 5/5 green. Distinguishing run on vnext master cddab84c: 0/5. pvk-reuse and pvk-parent Faulted with Parallel tasks produced conflicting output for key 'pvkSpawnChild', and the publish probes answered 200.
• Verified: OpenObserve shows the two same-order spawns running in parallel with status OK, and 0 "conflicting output" logs. Postgres shows three distinct child ids per parent, and the children are waiting.
• Full suite (FullyQualifiedName!~CrossDomainLab): 357/38/9 (404). All 29 distinct reds are also red on master, so there is no new red.
• TEST-SCENARIOS.md row added/updated: yes.

Notes

• Requires fix(tasks): keep parallel task responses apart with variableKey and a slot-aware merge vnext#1072. On a runtime without it, the scenario is red by design.
• Workflow definitions validate with variableKey on the current @burgan-tech/vnext-schema, because onExecuteTask has no additionalProperties: false. The field itself ships in feat(task): add optional variableKey to task entries vnext-schema#149.
• The children intentionally never finish: each run leaves pvk-child instances in waiting.

🤖 Generated with Claude Code

[coldtea-pr-lens]

Note

The title starts with test(parallel-variable-key-lab):, so PR Lens left this pull request undrawn. Comment @pr-lens draw to draw it

_{github.comment.notice: false in .github/pr-lens.yml turns this note off}

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

• Configuration used: defaults
• Review profile: CHILL
• Plan: Advanced
• Run ID: 514a8d3b-18c9-470c-a001-5d8665430ed2

• Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}