[ci-scan-feedback] Strengthen recently-closed KBE dedup and block JIT/GC workarounds

[github-actions]

Triggering signals

• Scanner duplicate filing ([ci-scan] Test failure: GitHub_19288 JIT assert genActualType mismatch in Morph under MinOpts stress #129028, rubric: filed 1 day after predecessor [ci-scan] JIT assertion: genActualType comparison mismatch in GitHub_19288 under MinOpts stress (recurrence) #129011 was closed as not_planned; recently-closed search should have caught it within 7-day window, #129028)
• Scanner duplicate filing ([ci-scan] Test failure: crossgen2_comparison.py asyncio incompatibility with Python 3.14 #129171, rubric: filed 4 days after predecessor [ci-scan] Build break: crossgen2_comparison.py fails with Python 3.14 asyncio.get_event_loop() removal #128992 was closed; recently-closed search missed it, #129171)
• Scanner duplicate rate: 6 duplicates out of 30 closures in 30d (20%) — [ci-scan] Test failure: ComInterfaceGenerator.Tests crashes with SIGBUS on NativeAOT linux-arm #128878, [ci-scan] Test failure: DelegateCombine1 segfault on NativeAOT linux-arm #128932, [ci-scan] GC stress assertion: CREATE_CHECK_STRING(pMT && pMT->Validate()) in object.cpp #128995, [ci-scan] Test failure: GC SanityCheck assertion under gcstress on windows x86 #129015, [ci-scan] Test failure: GitHub_19288 JIT assert genActualType mismatch in Morph under MinOpts stress #129028, [ci-scan] Test failure: crossgen2_comparison.py asyncio incompatibility with Python 3.14 #129171
• Fixer workaround rejection ([ci-fix] Needs review: Normalize -1 bufferSize in StreamReader/StreamWriter String ctors (refs #128062) #129360, MihaZupan: "I don't see a point in trying to workaround a codegen bug like this" — fixer opened a help-wanted PR changing StreamReader/StreamWriter buffer sizes to sidestep a JIT genActualType assertion, #129360)

Proposed edits

• create-kbe.instructions.md:95-110 — Expand the recently-closed KBE search from 2 query variations to 4: add a [ci-scan] title + test-name-stem search (catches title rewording like "JIT assert" vs "JIT assertion"), and a distinctive-assertion-substring body search (the assertion text is more stable than titles across scanner runs). Tied to the [ci-scan] Test failure: GitHub_19288 JIT assert genActualType mismatch in Morph under MinOpts stress #129028/[ci-scan] Test failure: crossgen2_comparison.py asyncio incompatibility with Python 3.14 #129171 duplicate signals.
• ci-failure-fix.md:182 — Clarify the JIT/GC fix-policy row: workarounds in unrelated code (e.g. changing library buffer sizes or API call patterns to sidestep a codegen bug) are equally out of bounds. Tied to the [ci-fix] Needs review: Normalize -1 bufferSize in StreamReader/StreamWriter String ctors (refs #128062) #129360 rejection signal.

Expected behavior change

The next scanner run will perform 4 search queries instead of 2 when checking for recently-closed KBE predecessors, catching cases where the predecessor title uses different phrasing for the same test/assertion. This should reduce the duplicate-closure rate (currently 20%). The next fixer run will skip workaround attempts for JIT/GC codegen bugs and go directly to a loop-in comment, avoiding rejected help-wanted PRs like #129360.

Note

🔒 Integrity filter blocked 1 item

The following item was blocked because it doesn't meet the GitHub integrity level.

• #128501 search_pull_requests: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".

To allow these resources, lower min-integrity in your GitHub frontmatter:

tools:
 github:
 min-integrity: approved # merged | approved | unapproved | none

Generated by CI Outer-Loop Failure Scanner — Feedback · ● 23.1M · ◷

---

Note

This was originally intended as a pull request, but the git push operation failed.

Workflow Run: View run details and download patch artifact

The patch file is available in the agent artifact in the workflow run linked above.

To create a pull request with the changes:

# Download the artifact from the workflow run
gh run download 27496184152 -n agent -D /tmp/agent-27496184152

# Create a new branch
git checkout -b ci-scan-feedback/strengthen-dedup-and-workaround-policy-bcc2b09de535bee9

# Apply the patch (--3way handles cross-repo patches where files may already exist)
git am --3way /tmp/agent-27496184152/aw-ci-scan-feedback-strengthen-dedup-and-workaround-policy.patch

# Push the branch to origin
git push origin ci-scan-feedback/strengthen-dedup-and-workaround-policy-bcc2b09de535bee9

# Create the pull request
gh pr create --title '[ci-scan-feedback] Strengthen recently-closed KBE dedup and block JIT/GC workarounds' --base main --head ci-scan-feedback/strengthen-dedup-and-workaround-policy-bcc2b09de535bee9 --repo dotnet/runtime

Show patch preview (77 of 77 lines)

From b818f520e0e09c6d357f95a1238ef6deb593e656 Mon Sep 17 00:00:00 2001
From: "github-actions[bot]" <github-actions[bot]@users.noreply.github.com>
Date: Sun, 14 Jun 2026 10:55:10 +0000
Subject: [PATCH] ci-scan-feedback: strengthen recently-closed dedup and block
 JIT/GC workarounds
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

- Expand recently-closed KBE search from 2 to 4 query variations
 (test-name stem, assertion substring) so title rewording between
 scanner runs no longer causes duplicate filings.
- Clarify fixer JIT/GC policy: workarounds in unrelated code
 (e.g. library buffer-size changes) are equally out of bounds.

Signals:
 Scanner: 6 duplicates in 30d (20% of closures). #129028 filed 1d
 after predecessor #129011 closed; #129171 filed 4d after #128992.
 Fixer: #129360 help-wanted PR rejected by MihaZupan — attempted
 StreamReader buffer-size workaround for a codegen bug.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---
 .github/workflows/ci-failure-fix.md | 2 +-
 .../shared/create-kbe.instructions.md | 22 ++++++++++++++-----
 2 files changed, 17 insertions(+), 7 deletions(-)

diff --git a/.github/workflows/ci-failure-fix.md b/.github/workflows/ci-failure-fix.md
index 25c99cc7..9326e5c9 100644
--- a/.github/workflows/ci-failure-fix.md
+++ b/.github/workflows/ci-failure-fix.md
@@ -179,7 +179,7 @@ Apply these fixer-specific bounds on top of the skill's guidance:
 | KBE pipeline / area | Fix policy |
 |---|---|
 | Mobile (ios/tvos/maccatalyst/android/wasm/wasi) | Small test/csproj/condition fixes in bounds are fair game. |
-| JIT / GC / PGO stress (codegen) | JIT/GC product fixes are OUT of bounds for any PR — no safe diff is producible, so loop in JIT/GC owners with a comment. |
+| JIT / GC / PGO stress (codegen) | JIT/GC product fixes are OUT of bounds for any PR — no safe diff is producible, so loop in JIT/GC owners with a comment. **Workarounds in unrelated c
... (truncated)

[kotlarmilos]

Superseded by #129418. The 2026-06-15 ci-scan-feedback tick carries forward the same proposed edits — expanding the recently-closed KBE search from 2 to 4 query variations in create-kbe.instructions.md and blocking JIT/GC workarounds — along with this tick’s un-pushed patch, and it covers the same triggering signals (#129171, #129028, #129360).

Closing in favor of #129418 to avoid a duplicate feedback PR.