merge-wave

name	merge-wave
description	Triage N open PRs with sharded test262 results, rank by net pass delta, merge winners, close catastrophes. Use when multiple PRs have completed CI and you want to land a batch.

Merge Wave Protocol

Triage every open PR that has finished sharded test262 CI, compute its real net delta, and execute a ranked merge order in one pass.

When to use

• Multiple PRs ready to merge after a morning of dev work
• After a sprint's worth of issues finished CI at roughly the same time
• When you want to cross a pass-rate milestone in one batch

Step 1: Gather candidate PRs

gh pr list --limit 20 --json number,title,headRefName,mergeable,mergeStateStatus

Filter to PRs with mergeStateStatus: UNSTABLE | CLEAN | BEHIND — skip DIRTY (conflict, ping dev to refresh) and DRAFT.

Step 2: Download every PR's merged-report artifact

The sharded workflow uploads test262-merged-report per run. Find each PR's run ID via the merge shard reports check:

mkdir -p output/pr-artifacts
for pr in <LIST>; do
  run_id=$(gh pr view $pr --json statusCheckRollup --jq '.statusCheckRollup[] | select(.name=="merge shard reports") | .detailsUrl' | grep -oE 'runs/[0-9]+' | head -1 | cut -d/ -f2)
  mkdir -p output/pr-artifacts/pr-$pr
  gh run download $run_id -n test262-merged-report -D output/pr-artifacts/pr-$pr >/dev/null 2>&1
done

Artifact filenames: test262-report-merged.json, test262-regressions.txt, test262-results-merged.jsonl.

Step 3: Compute and rank net deltas

BASELINE=$(python3 -c "import json; d=json.load(open('benchmarks/results/test262-current.json')); print(d['summary']['pass'])")

for pr in <LIST>; do
  rpt=output/pr-artifacts/pr-$pr/test262-report-merged.json
  [ -f "$rpt" ] && python3 -c "
import json
d=json.load(open('$rpt'))
s=d['summary']
delta=s['pass']-$BASELINE
print(f'PR #$pr  pass={s[\"pass\"]}  delta={delta:+d}  ce={s[\"compile_error\"]}')
"
done | sort -t= -k3 -n -r

Step 4: Classify each PR

Thresholds (tune per context):

Delta	Action
< −500	CATASTROPHIC → close immediately with diagnosis. Do NOT merge.
−500 to −20	Real regression → close, open follow-up issue, ask dev to investigate root cause
−20 to +20	MARGINAL → check if regressions are false positives; if yes merge, if no close
+20 to +100	GOOD → sample regressions, admin-merge if regressions are false positives or acceptable
+100+	STRONG → admin-merge, flag regressions in a follow-up issue

Sample regressions for any PR where the regression count is > 5% of the delta:

head -40 output/pr-artifacts/pr-<NUM>/test262-regressions.txt

Look for known false-positive patterns (see feedback_regression_analysis.md): harness wrapper coincidence, String.prototype wrapper tests, test expecting SyntaxError where we throw a different error at a matching phase, etc.

Step 5: Execute the ranked merges

Descending by net delta, skipping catastrophes and conflicts:

gh pr merge <NUM> --merge --admin --body "Net +<DELTA> pass. <CLASSIFICATION NOTES>."

Use --admin when mergeStateStatus: UNSTABLE (sharded CI flagged regressions but the net is positive). Use plain --merge (no --admin) when mergeStateStatus: CLEAN.

Step 6: Close losers

gh pr close <NUM> --comment "Net <DELTA> pass. <ROOT CAUSE IF KNOWN>. Reopening issue for a narrower retry."

Step 7: Refresh ALL open PR branches (required after every wave)

After each merge, main advances. Every open PR branch needs origin/main merged in so the quality gate ("Verify origin/main is merged into branch") stays green. Do this for ALL open branches, not just DIRTY ones:

gh pr list --state open --json number,headRefName --jq '.[].headRefName' | while read branch; do
  wt="/workspace/.claude/worktrees/$branch"
  if [ -d "$wt" ]; then
    git -C "$wt" fetch origin main
    git -C "$wt" merge origin/main --no-edit \
      && git -C "$wt" push origin "$branch" \
      && echo "✓ refreshed $branch" \
      || echo "⚠ CONFLICT on $branch"
  else
    echo "→ no local worktree for $branch — send dev a refresh request"
  fi
done

For every DIRTY (conflicting) PR with a good pre-wave delta, also SendMessage to the assigned dev:

"PR #NNN conflicts with new main after today's merges. Please git merge origin/main in your worktree, resolve, push. Your pre-merge delta was +NNN."

Step 8: Verify and log

git fetch origin main --quiet
git log --oneline origin/main -<NUM_MERGES+2>

Append to plan/diary.md:

• PRs merged with deltas
• PRs closed with reasons
• Follow-up issues filed
• Total net delta
• Gap to next milestone

Output

A short summary to the user:

Merged: #43 (+258), #68 (+106), #71 (+56), #64 (+34)
Closed: #72 (catastrophic -18504), #75 (-114), #65 (orphaned +2)
DIRTY: #74 (dev-1016 refreshing)
Net: +454 pass. Baseline 20711 → ~21165. Gap to 50%: ~400.

Notes

• Baseline drift during merge: the deltas you computed are vs the old baseline. Each merge adds its pass delta but also potentially introduces new regressions that weren't visible in per-PR CI. The final sharded refresh commit is the source of truth.
• Skip-CI tag on baseline refresh: baseline commits include [skip ci] to avoid retriggering the sharded workflow. PR #70 added an explicit Pages deploy dispatch so the landing page still updates.
• Allow-regressions override: if the post-merge sharded run fails on regressions, re-run with gh workflow run "Test262 Sharded" --ref main -f allow_regressions=true to force the baseline refresh.