⚡
Evaluation in progress...
Cancel
⚠️
A previous evaluation was interrupted.
Resume
Agent Builder
Discovery
Evaluation
Agent Evaluation
Generate and run evaluation question sets against any deployed WXO agent
1
Data Source — Discovery Report
Select a saved report to generate questions
▼
Saved Discovery Reports
— loading reports… —
Report loaded
2
WXO Connection & Agent
Connect and select an agent
▼
WXO URL
IBM Cloud API Key
Connect
Select Agent
Agent connected
3
Evaluation Questions
Generate or load a question set
▼
Load Saved Question Set
— select saved set —
Load
Delete
Upload QA CSV
Drop a CSV file here or click to browse
Format: id, question, expected_answer, category
LLM Provider
OpenRouter
Watsonx.ai
Model
OpenRouter API Key
Generate Questions
Generated Questions
+ Add Custom
data_query
tool_specific
out_of_scope
corner_case
Add
Questions
All
None
Type
Question
Expected Answer
Save Set
4
Run & Results
Execute evaluation and score with LLM-as-judge
▼
Judge LLM
OpenRouter
Watsonx.ai
Model
OpenRouter API Key
Timeout (sec)
Run Selected
Run All
⏹ Cancel
Score with Judge
Starting...
HTML Report
Export JSON
Load Past Runs