Tactic: Evidence gathering, cross-examination, and quality judgment. External evidence is collected, presented, challenged, and scored for relevance and reliability.
SkillScore91Maintained
Automatic
◆Research pack
Last verified 8/22/2026
What it changes
Passed 20 of 22 test cases with the skill, and 5 of 22 without it.
Without skill23%
With skill91%
Checks it passed that the model alone missed
Uses 0.0 to 1.0 numerical scales for evidence scoring.
Delivers a final tournament bracket result from the judge.
Uses a 0.0 to 1.0 decimal scale for evidence quality metrics.
and 20 more in the full report
15 fixed5 passed either way2 still failing0 broke
Pass-rate lift
+68 pp
Tokens
-2%
Agent turns
0%
Response time
-24%
Model gemini-3.6-flashMeasured 8/22/2026v120 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 15 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 15
Prompt
We plan to gather evidence regarding whether migrating our monolith to serverless functions reduces infrastructure overhead. Model guidance often suggests uncapped searching or arbitrary caps like 50 queries; what are the allowed search budget limit options when configuring evidence gathering agents?
What was checked, without and with the skill
Checked
Without skill
With skill
Restricts evidence search budget choices to 2, 5, or 10 searches.
Missed
Passed
case-04 · gemini-3.6-flash · 8/22/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.