fix(hitl): replace vestigial 'HITL result' natural-language in quality_02 prompt#547
Conversation
…y_02 prompt PR #544 bot review nit: quality_02_result_downstream.yaml line 27 still said "reads the reviewer's decision from the HITL result" while line 29 already says "the HITL output". Replace for full consistency. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
Claude finished @tmatup's task in 54s —— View job Review:
|
| Area | Status | Notes |
|---|---|---|
| Frontmatter | N/A | No frontmatter changes |
| E2E Tests | OK | Test file content updated, structure intact |
| Skill Body | N/A | No skill body changes |
| References & Assets | N/A | No reference/asset changes |
| Repo Hygiene | OK | Scoped change, no secrets, no cross-skill refs |
Issues for Manual Review
None found.
Conclusion
Approve. The one-word fix is correct and makes quality_02_result_downstream.yaml internally consistent. The only observation is that three other test files have the same vestigial "HITL result" phrasing — worth fixing in this PR or a follow-up.
…tests Bot review on #547 caught three more instances of the same vestigial "HITL result" wording in adjacent maestro-flow HITL test prompts that the original cherry-pick missed: - quality_03_boolean_decision.yaml:28 — "from the HITL result:" - smoke_03_multi_outcome_routing.yaml:26 — "from the HITL result" - smoke_03_multi_outcome_routing.yaml:33 — "reference the HITL result using the" All three rewritten to "HITL output" to match the v1.0 schema rename. The smoke_03:33 occurrence was particularly clear — the very next line already uses `$vars.<nodeId>.output.<fieldName>` while the prose said "HITL result", so the prompt was internally inconsistent. `grep -rn "HITL result" tests/tasks/uipath-maestro-flow/hitl/` now returns no matches. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Review Response1. Same "HITL result" phrasing persists in two sibling test files — Fixed in f547a9a. Verified all three flagged occurrences and swept them in. The smoke_03:33 case was especially clear — the very next line already uses PR title still accurately describes the intent (vestigial "HITL result" → "HITL output" cleanup); now does it as a complete sweep across the four files in 🤖 Generated with Claude Code Co-Authored-By: Claude noreply@anthropic.com |
Summary
tests/tasks/uipath-maestro-flow/hitl/quality_02_result_downstream.yamlsays "reads the reviewer's decision from the HITL result" while line 28 already says "the HITL output". After the v1.0 schema rename (#528 / #529 —.result→.output), the natural-language descriptions in test prompts should also use "output" consistently. This is a follow-up to a bot review nit on closed PR fix(hitl): align maestro-flow HITL docs + tests to v1.0 schema (.output not .result) #544.Background
PR #544 (
fix(hitl): align maestro-flow HITL docs + tests to v1.0 schema (.output not .result)) overlapped substantially with #528 and was closed when #528 landed. However, the bot review on #544 caught one additional inconsistency that #528 didn't fix — line 27 ofquality_02_result_downstream.yaml. This PR carries just that one-word fix forward onto current main.Test plan
result→outputon line 27 of the test prompt.🤖 Generated with Claude Code