Setting the file. One moment.
Subchapter 14.1
tests/eval.yaml
YAML64 lines2 KB
name: waza-interactive-triggers
skill: waza-interactive
version: "1.0"
executor: mock
tasks:
- id: trigger-run-evals
name: "Trigger: run my evals"
prompt: "Can you run my evals and tell me how my skill is doing?"
expected: "Should activate waza-interactive skill and start Scenario 2 (Run and Interpret Results)"
validators:
- type: text
config:
regex_match:
- "waza_eval_run|run.*eval|Scenario 2"
- id: trigger-create-eval
name: "Trigger: create eval suite"
prompt: "I need to create an eval suite for my new code-reviewer skill"
expected: "Should activate waza-interactive skill and start Scenario 1 (Create a New Eval)"
validators:
- type: text
config:
regex_match:
- "waza_eval_list|waza init|Scenario 1|scaffold"
- id: trigger-compare-models
name: "Trigger: compare models"
prompt: "Which model works better for my skill — gpt-4o or claude-sonnet?"
expected: "Should activate waza-interactive skill and start Scenario 3 (Compare Models)"
validators:
- type: text
config:
regex_match:
- "compare|model.*A|model.*B|Scenario 3"
- id: trigger-debug-failing
name: "Trigger: debug failing tests"
prompt: "My skill keeps failing its eval tests, can you help me debug?"
expected: "Should activate waza-interactive skill and start Scenario 4 (Debug a Failing Skill)"
validators:
- type: text
config:
regex_match:
- "waza_skill_check|debug|fail|Scenario 4"
- id: trigger-ship-readiness
name: "Trigger: is my skill ready"
prompt: "Is my skill ready to ship? Give me the full readiness check."
expected: "Should activate waza-interactive skill and start Scenario 5 (Ship Readiness Check)"
validators:
- type: text
config:
regex_match:
- "readiness|checklist|waza_skill_check|Scenario 5"
- id: anti-trigger-general-coding
name: "Anti-trigger: general coding request"
prompt: "Write me a Python function that sorts a list"
expected: "Should NOT activate waza-interactive skill — this is general coding"
validators:
- type: text
config:
regex_match:
- "def sort|sorted|python"