Telemetry from ai_reviews
Explore risk patterns, rerun behavior, and review outcomes for your filtered pull requests.
Start here for a quick health snapshot, then use the grouped sections below to inspect root causes and trends.
Avg PR risk
27.9
Re-review rate
50%
Avg runs / PR
1.50
Eventual approval
100%
Automatically detected patterns that may need triage or process adjustments.
Vulnerability hotspot: Repos with critical + high vulnerabilities
2Distribution, trend, and correlation charts to explain what drives risk and blocking outcomes.
Read this section from top to bottom: transition patterns, cohort progression, and remaining blockers by rerun index.
Most common verdict changes between consecutive runs for the same PR.
| From | To | Count |
|---|---|---|
| CHANGES_REQUESTED | APPROVED | 1 |
Tracks the share of PRs approved on run 1, run 2, or run 3+ by first run week.
| Cohort week | Approved by run1 | Approved by run2 | Approved by run3+ |
|---|---|---|---|
| 2026-09-07 | 1 | 1 | 0 |
Percentage of PRs that remain blocking after each rerun index.
| Run index | Still blocking | Observed PRs | Still blocking % |
|---|---|---|---|
| Run 1 | 2 | 2 | 100% |
| Run 2 | 1 | 1 | 100% |
Sorted by highest risk. At most 25 rows; scoped to your current filters.
No pull requests in this window scored 70 or above. Try widening the date range or relaxing other filters.