From 7632a7796ce216b71129ec52f304fb906581750d Mon Sep 17 00:00:00 2001 From: mrizzi Date: Fri, 10 Jul 2026 17:51:16 +0200 Subject: [PATCH] fix(verify-pr): add explicit Issue Type field to Step 6d sub-task creation templates Add `**Issue Type:** Sub-task` as the first field in all three jira.create_issue templates in Step 6d (review feedback, CI failure, and eval failure sub-tasks). Without this field, the LLM nondeterministically infers the issue type, causing ~40% of sub-tasks to be created as standalone Tasks instead of Sub-tasks with proper parent-child hierarchy. Add eval assertion to verify-pr eval 3 to validate that sub-task creation output specifies Issue Type as Sub-task. Implements TC-5181 Assisted-by: Claude Code --- evals/verify-pr/evals.json | 3 ++- plugins/sdlc-workflow/skills/verify-pr/SKILL.md | 3 +++ 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/evals/verify-pr/evals.json b/evals/verify-pr/evals.json index 98e7f846..94fa9608 100644 --- a/evals/verify-pr/evals.json +++ b/evals/verify-pr/evals.json @@ -59,7 +59,8 @@ "The report contains a Test Change Classification row with ADDITIVE — only new test files were added (tests/api/sbom_delete.rs is a new file)", "Convention upgrade eligibility is evaluated for review comment 30002 (index suggestion) — the review classification output (review-30002.md) or the report's Style/Conventions analysis explains whether the suggestion matches a documented or demonstrated project convention", "Review comment 30002 (index suggestion) does NOT result in a sub-task — the suggestion classification is correct (suggestive language, no directive) and no project convention in the fixture data backs an upgrade from suggestion to code change request", - "Eval Quality is N/A because no eval result reviews exist in the PR — the 3-criteria detection (author github-actions[bot], marker ## Eval Results, footer sdlc-workflow/run-evals) found no matches, so Eval Quality does not affect the Test Quality combination" + "Eval Quality is N/A because no eval result reviews exist in the PR — the 3-criteria detection (author github-actions[bot], marker ## Eval Results, footer sdlc-workflow/run-evals) found no matches, so Eval Quality does not affect the Test Quality combination", + "The sub-task creation for comment 30001 explicitly specifies Issue Type as Sub-task — the subtask-30001.md file, the report's sub-task section, or the review classification output (review-30001.md) indicates the Jira issue is created with issueTypeName Sub-task (not as a standalone Task) to ensure parent-child hierarchy with the parent task" ] }, { diff --git a/plugins/sdlc-workflow/skills/verify-pr/SKILL.md b/plugins/sdlc-workflow/skills/verify-pr/SKILL.md index 2a06b114..3069fad7 100644 --- a/plugins/sdlc-workflow/skills/verify-pr/SKILL.md +++ b/plugins/sdlc-workflow/skills/verify-pr/SKILL.md @@ -485,6 +485,7 @@ For each code change request (including upgraded suggestions) that requires a fi create a Jira sub-task: jira.create_issue with: +- **Issue Type:** Sub-task - **Parent:** the current task's Jira issue ID - **Summary:** concise description of the required fix - **Labels:** `["ai-generated-jira", "review-feedback"]` @@ -514,6 +515,7 @@ Process `create-sub-task` actions from the Correctness sub-agent. For each actio files, and Root cause fields: jira.create_issue with: + - **Issue Type:** Sub-task - **Parent:** the current task's Jira issue ID - **Summary:** the action's Title (e.g., "Fix failing lint check: unused import in handler.rs") - **Labels:** `["ai-generated-jira", "review-feedback"]` @@ -556,6 +558,7 @@ and sub-task creation below. 3. **Create sub-task:** For each failing eval, create a Jira sub-task: jira.create_issue with: + - **Issue Type:** Sub-task - **Parent:** the current task's Jira issue ID - **Summary:** "Fix assertion failures: " - **Labels:** `["ai-generated-jira", "eval-failure"]`