Passed 17 of 20 test cases with the skill, and 10 of 20 without it.
Without skill52%
With skill87%
Checks it passed that the model alone missed
Prohibits POST operations.
Logs inspection steps in GAIT.
Restricts operations to GET requests.
and 10 more in the full report
8 fixed12 passed either way3 still failing0 broke
Pass-rate lift
+35 pp
Tokens
+263%
Agent turns
0%
Response time
-42%
Model gemini-3.6-flashMeasured 8/16/2026v120 of 23 cases passed with skill · 20 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 8 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 8
Prompt
We want to run a pyATS CLI command to show virtual server status on device bigip-01 using pyats_run_show_command. How should the invocation command string payload be formatted?
What was checked, without and with the skill
Checked
Without skill
With skill
Uses pyats_run_show_command in the CLI invocation.
Missed
Passed
Includes show ltm virtual in the command string payload.
Passed
Passed
case-19 · gemini-3.6-flash · 8/16/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.