How the reviewer is doing
24 reviews: 20 finished, 3 failed, 1 too large.
Where findings go
33
Proposed
What the reviewer handed in.
6
Refuted
Dropped by the second look: the code showed they were wrong.
27
Kept
Stored, and listed in the Check Run.
12
Posted
Commented on the pull request. 2 more were already there from an earlier push.
0
Marked wrong
Given a 👎 by someone on GitHub.
Reviews per day, last 30 days
Show as a table
| Day | Reviews | Findings | Posted |
|---|---|---|---|
| 12 Nov | 3 | 3 | 2 |
The second look
Each finding checked again against the code, before anything is posted.
- Confirmed
- 5
- Unsure, shown but not posted
- 0
- No verdict
- 0
- Refuted and dropped
- 6
- Refuted, of all checked
- 55%
What people thought
👍 and 👎 on the comments the reviewer posted.
- Comments on GitHub
- 12
- With a reaction
- 0 (0%)
- 👍 at least once
- 0
- 👎 at least once
- 0
- Resolved by the reviewer
- 1
- Resolved by a person
- 0
The agent
Reviews an AI model answered.
- Reviews
- 10
- Average time
- 35s
- Average turns
- 3.2
- Tokens in / out
- 103k / 10k
- Cost
- $0.00