Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →juliusbrussee/investigate-first
Diagnose ambiguous failures before editing.
Passed 5 of 13 test cases with the skill, and 7 of 13 without it.
2 fixed8 passed either way10 still failing5 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 2 test cases the skill fixed. Same model, same prompt, with and without the skill.
A customer reported that user.is_active remains false after clicking the verification link. We inspected the DB write logs and verified that VerificationToken.consume() sets token.used = true but fails to update user.is_active due to a missing transaction commit on line 42. Do we need to inspect the notification service and front-end state management as well before reporting our findings?
| Checked | Without skill | With skill |
|---|---|---|
| stops further exploration once the missing transaction commit is identified as the cause | Missed | Passed |
| reports the missing transaction commit as the cause | Passed | Passed |
| refrains from modifying product code | Passed | Passed |
The instructions your model reads when the skill loads.
Gather evidence before changing product code.
Report cause and proof. Make no fix unless task authorizes implementation.
test cases passed, vs 7 / 13 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.