review-pr

aiter PR Review

Step 1 — Fetch

PR=$1  # PR number from skill argument
REPO="ROCm/aiter"

# Full metadata
gh pr view $PR --repo $REPO --json title,body,number,labels,files,author,reviews,comments > /tmp/pr_meta.json

# Diff
gh pr diff $PR --repo $REPO > /tmp/pr.diff

# Linked issue (extract from body "fix: #NNN" or "close #NNN")
ISSUE=$(cat /tmp/pr_meta.json | python3 -c "
import json,re,sys
body = json.load(sys.stdin).get('body','') or ''
m = re.search(r'(?:fix|close|resolve)[s]?[: ]*#(\d+)', body, re.I)
print(m.group(1) if m else '')
")
[ -n "$ISSUE" ] && gh issue view $ISSUE --repo $REPO --json title,body > /tmp/pr_issue.json

# Prior reviewer comments (top-level)
cat /tmp/pr_meta.json | python3 -c "
import json,sys
d = json.load(sys.stdin)
for r in d.get('reviews',[]):
    b = (r.get('body','') or '').strip()
    if b: print(f'[REVIEW {r[\"author\"][\"login\"]}] {b[:200]}')
for c in d.get('comments',[]):
    b = (c.get('body','') or '').strip()
    if b: print(f'[COMMENT {c[\"author\"][\"login\"]}] {b[:200]}')
"

# Inline review comments (line-level code comments — often more specific than top-level)
gh api "repos/$REPO/pulls/$PR/comments" | python3 -c "
import json,sys
comments = json.load(sys.stdin)
for c in comments:
    author = c.get('user',{}).get('login','')
    body = (c.get('body','') or '').strip()
    path = c.get('path','')
    line = c.get('line') or c.get('original_line','')
    if body and 'copilot' not in author.lower() and 'bot' not in author.lower():
        print(f'[INLINE {author}] {path}:{line}')
        print(f'  {body[:250]}')
" 2>/dev/null

Read the diff and PR body before proceeding.

Cross-file verification — do this before reporting any kernel/dispatch finding. The diff shows changed lines, not the whole story. Grep the entire symbol family, not just files in the diff:

• Sync/fence/atomics, or the "other half" of a scatter, may live in a .cuh/.h rather than the .cu in the diff. Grep .cu + .cuh + .h together before claiming "no synchronization" or "no bounds check." (aiter#3802: a "kernel has no sync" finding was a false positive — the signal barrier was in the .cuh.)
• Read the actual function in the head file, not just the diff hunk, before claiming a branch/else is missing. (aiter#4098: a "compares raw uint8 vs float" finding was false — the reader had a conditional maybe_view_fp8() the diff never showed.)