Files
promptfoo--promptfoo/test/fixtures/file-script-assertions/comprehensive-llm-rubric-test.yaml
T
wehub-resource-sync 0d3cb498a3
CI / Shell Format Check (push) Has been cancelled
CI / Check Ruby (3.4) (push) Has been cancelled
CI / CI Config (push) Has been cancelled
CI / Test on Node ${{ matrix.node }} and ${{ matrix.os }}${{ matrix.shard && format(' (shard {0}/3)', matrix.shard) || '' }} (push) Has been cancelled
CI / Build on Node ${{ matrix.node }} (push) Has been cancelled
CI / Style Check (push) Has been cancelled
CI / Generate Assets (push) Has been cancelled
CI / Check Python (3.14) (push) Has been cancelled
CI / Check Python (3.9) (push) Has been cancelled
CI / Build Docs (push) Has been cancelled
CI / Code Scan Action (push) Has been cancelled
CI / Site tests (push) Has been cancelled
CI / webui tests (push) Has been cancelled
CI / Run Integration Tests (push) Has been cancelled
CI / Run Smoke Tests (push) Has been cancelled
CI / Go Tests (push) Has been cancelled
CI / Share Test (push) Has been cancelled
CI / Redteam (Production API) (push) Has been cancelled
CI / Redteam (Staging API) (push) Has been cancelled
CI / GitHub Actions Lint (push) Has been cancelled
CI / Check Ruby (3.0) (push) Has been cancelled
release-please / release-please (push) Has been cancelled
release-please / build (push) Has been cancelled
release-please / publish-npm (push) Has been cancelled
release-please / publish-npm-backfill (push) Has been cancelled
release-please / docker (push) Has been cancelled
release-please / publish-code-scan-action (push) Has been cancelled
release-please / attest-code-scan-action (push) Has been cancelled
Deploy local.promptfoo.app / Deploy to Cloudflare Pages (push) Has been cancelled
Test and Publish Multi-arch Docker Image / test (push) Has been cancelled
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-amd64 platform:linux/amd64 runner:ubuntu-latest]) (push) Has been cancelled
Test and Publish Multi-arch Docker Image / build-docker-and-push-digests (map[digest-suffix:linux-arm64 platform:linux/arm64 runner:ubuntu-24.04-arm]) (push) Has been cancelled
Test and Publish Multi-arch Docker Image / merge-docker-digests (push) Has been cancelled
Test and Publish Multi-arch Docker Image / Attest Multi-arch Image (push) Has been cancelled
Validate Renovate Config / Validate Renovate Configuration (push) Has been cancelled
chore: import upstream snapshot with attribution
2026-07-13 13:24:08 +08:00

301 lines
8.6 KiB
YAML

# Comprehensive llm-rubric and normal usage test for #6200
# Tests both file:// scripts AND standard usage patterns
providers:
- id: echo
prompts:
- '{{prompt}}'
defaultTest:
options:
provider: echo
tests:
# ===========================================
# STANDARD USAGE (no file:// - must not break)
# ===========================================
# Test 1: Standard llm-rubric with direct string value
- description: 'llm-rubric with direct string value (standard usage)'
vars:
prompt: 'The capital of France is Paris'
assert:
- type: llm-rubric
value: 'Check that the response correctly identifies Paris as the capital of France'
# Test 2: Standard llm-rubric with simple criteria
- description: 'llm-rubric with simple criteria'
vars:
prompt: 'Python is a programming language'
assert:
- type: llm-rubric
value: 'The response should mention Python'
# Test 3: Standard contains with direct value
- description: 'contains with direct string (standard usage)'
vars:
prompt: 'Hello world'
assert:
- type: contains
value: 'Hello'
# Test 4: Standard equals with direct value
- description: 'equals with direct string (standard usage)'
vars:
prompt: 'exact match'
assert:
- type: equals
value: 'exact match'
# Test 5: Standard regex with direct pattern
- description: 'regex with direct pattern (standard usage)'
vars:
prompt: 'Order #12345'
assert:
- type: regex
value: "Order #\\d+"
# Test 6: Standard similar with direct value
- description: 'similar with direct value (standard usage)'
vars:
prompt: 'The weather is nice today'
assert:
- type: similar
value: 'The weather is nice today'
threshold: 0.9
# Test 7: Standard starts-with
- description: 'starts-with with direct value'
vars:
prompt: 'Hello, how are you?'
assert:
- type: starts-with
value: 'Hello'
# Test 8: Standard contains-all with array
- description: 'contains-all with direct array (standard usage)'
vars:
prompt: 'I love apples and bananas'
assert:
- type: contains-all
value:
- apples
- bananas
# Test 9: Standard contains-any
- description: 'contains-any with direct array'
vars:
prompt: 'I prefer oranges'
assert:
- type: contains-any
value:
- apples
- oranges
- bananas
# ===========================================
# LLM-RUBRIC WITH FILE:// SCRIPTS (the fix)
# ===========================================
# Test 10: llm-rubric with JS file:// - simple
- description: 'llm-rubric with JS script returning simple string'
vars:
prompt: 'SCRIPT_OUTPUT_12345'
assert:
- type: llm-rubric
value: file://rubric-generator.cjs:knownValue
# Test 11: llm-rubric with JS file:// - dynamic based on context
- description: 'llm-rubric with JS script using context vars'
vars:
prompt: 'Machine learning is a subset of AI'
topic: 'machine learning'
assert:
- type: llm-rubric
value: file://rubric-generator.cjs:rubric
# Test 12: llm-rubric with JS file:// - object return
- description: 'llm-rubric with JS script returning object'
vars:
prompt: 'Test response'
assert:
- type: llm-rubric
value: file://rubric-generator.cjs:rubricObject
# Test 13: llm-rubric with Python file://
- description: 'llm-rubric with Python script'
vars:
prompt: 'Neural networks explanation'
topic: 'neural networks'
assert:
- type: llm-rubric
value: file://rubric-generator.py:rubric
# Test 14: llm-rubric with Ruby file://
- description: 'llm-rubric with Ruby script'
vars:
prompt: 'Deep learning concepts'
topic: 'deep learning'
assert:
- type: llm-rubric
value: file://rubric-generator.rb:rubric
# ===========================================
# MIXED SCENARIOS
# ===========================================
# Test 15: Multiple assertions - mix of file:// and direct
- description: 'Mixed assertions - file:// and direct values'
vars:
prompt: 'SCRIPT_OUTPUT_12345 is the answer'
assert:
- type: contains
value: file://rubric-generator.cjs:knownValue
- type: contains
value: 'answer'
- type: starts-with
value: file://rubric-generator.cjs:knownValue
# Test 16: llm-rubric alongside other assertions
- description: 'llm-rubric with other assertion types'
vars:
prompt: 'The result is SCRIPT_OUTPUT_12345'
assert:
- type: llm-rubric
value: file://rubric-generator.cjs:knownValue
- type: contains
value: 'result'
- type: regex
value: "SCRIPT_OUTPUT_\\d+"
# ===========================================
# OTHER ASSERTIONS WITH FILE://
# ===========================================
# Test 17: equals with file://
- description: 'equals with JS file:// script'
vars:
prompt: 'SCRIPT_OUTPUT_12345'
assert:
- type: equals
value: file://rubric-generator.cjs:knownValue
# Test 18: contains with file://
- description: 'contains with JS file:// script'
vars:
prompt: 'Result: SCRIPT_OUTPUT_12345'
assert:
- type: contains
value: file://rubric-generator.cjs:knownValue
# Test 19: regex with file:// (dynamic pattern)
- description: 'regex with file:// script generating pattern'
vars:
prompt: 'Code: 99999'
pattern: "\\d{5}"
assert:
- type: regex
value: file://rubric-generator.cjs:getPattern
# Test 20: starts-with with file://
- description: 'starts-with with file:// script'
vars:
prompt: 'SCRIPT_OUTPUT_12345 begins here'
assert:
- type: starts-with
value: file://rubric-generator.cjs:knownValue
# Test 21: icontains with file://
- description: 'icontains (case insensitive) with file:// script'
vars:
prompt: 'script_output_12345 lowercase'
assert:
- type: icontains
value: file://rubric-generator.cjs:knownValue
# Test 22: not-contains with file://
- description: 'not-contains with file:// script'
vars:
prompt: 'This has something else'
assert:
- type: not-contains
value: file://rubric-generator.cjs:knownValue
# Test 23: contains-all with file:// returning array
- description: 'contains-all with file:// array return'
vars:
prompt: 'Has reference one and reference two'
assert:
- type: contains-all
value: file://rubric-generator.cjs:referenceArray
# Test 24: contains with file:// numeric return
- description: 'contains with file:// numeric return'
vars:
prompt: 'The answer is 0'
assert:
- type: contains
value: file://rubric-generator.cjs:numericValue
# ===========================================
# EDGE CASES
# ===========================================
# Test 25: Empty string from script
- description: 'equals empty string from file:// script'
vars:
prompt: ''
assert:
- type: equals
value: file://rubric-generator.cjs:emptyValue
# Test 26: Nunjucks template in direct value
- description: 'llm-rubric with nunjucks template'
vars:
prompt: 'Answer about Paris'
city: 'Paris'
assert:
- type: llm-rubric
value: 'Check that the response mentions {{city}}'
# Test 27: Multiple llm-rubric assertions
- description: 'Multiple llm-rubric assertions'
vars:
prompt: 'Comprehensive answer about AI'
assert:
- type: llm-rubric
value: 'The response should be about AI'
- type: llm-rubric
value: 'The response should be clear and concise'
# ===========================================
# JAVASCRIPT ASSERTION REGRESSION
# ===========================================
# Test 28: javascript assertion with file:// (uses return as result)
- description: 'javascript assertion file:// (regression test)'
vars:
prompt: 'contains expected keyword'
assert:
- type: javascript
value: file://rubric-generator.cjs:gradingFunction
# Test 29: javascript assertion inline (regression test)
- description: 'javascript assertion inline (regression test)'
vars:
prompt: 'hello world'
assert:
- type: javascript
value: output.includes("hello")
# Test 30: javascript assertion with complex logic
- description: 'javascript assertion with complex inline logic'
vars:
prompt: 'The answer is 42'
assert:
- type: javascript
value: |
const num = output.match(/\d+/);
return num && parseInt(num[0]) === 42;