Skip to content
GitHubDiscord

Overview

Complete API documentation for Giskard Checks β€” a comprehensive testing framework for AI applications.

from giskard.checks import (
# Core types
Check,
CheckResult,
CheckStatus,
Interaction,
Trace,
InteractionSpec,
Scenario,
TestCase,
# Built-in checks
FnCheck,
StringMatching,
Equals,
NotEquals,
LesserThan,
GreaterThan,
LesserThanEquals,
GreaterEquals,
# Configuration
set_default_generator,
get_default_generator,
)
# LLM-based checks
from giskard.checks import (
BaseLLMCheck,
LLMCheckResult,
Groundedness,
Conformity,
LLMJudge,
SemanticSimilarity,
)
giskard.checks/
β”œβ”€β”€ core/ # Base classes and fundamental types
β”‚ β”œβ”€β”€ check.py # Check base class
β”‚ β”œβ”€β”€ result.py # CheckResult, CheckStatus
β”‚ β”œβ”€β”€ scenario.py # Scenario
β”‚ β”œβ”€β”€ testcase.py # TestCase
β”‚ β”œβ”€β”€ extraction.py # Extractors
β”‚ └── interaction/ # Interaction types
β”œβ”€β”€ builtin/ # Ready-to-use checks
β”‚ β”œβ”€β”€ fn.py # FnCheck
β”‚ β”œβ”€β”€ text_matching.py # StringMatching, RegexMatching
β”‚ β”œβ”€β”€ comparison.py # Equals, NotEquals, etc.
β”‚ └── semantic_similarity.py
β”œβ”€β”€ judges/ # LLM-based checks
β”‚ β”œβ”€β”€ base.py # BaseLLMCheck, LLMCheckResult
β”‚ β”œβ”€β”€ groundedness.py # Groundedness
β”‚ β”œβ”€β”€ conformity.py # Conformity
β”‚ └── judge.py # LLMJudge
β”œβ”€β”€ scenarios/ # Multi-step workflows
β”‚ β”œβ”€β”€ builder.py # Scenario builder
β”‚ β”œβ”€β”€ runner.py # ScenarioRunner
β”‚ └── suite.py # Suite
β”œβ”€β”€ testing/ # Testing utilities
β”‚ β”œβ”€β”€ runner.py # TestCaseRunner
β”‚ └── spy.py # WithSpy
β”œβ”€β”€ generators/ # Input generators
β”‚ └── user.py # UserSimulator
└── utils/ # Helper utilities
β”œβ”€β”€ normalization.py
β”œβ”€β”€ value_provider.py
β”œβ”€β”€ parameter_injection.py
└── generator.py