behavioral-testing 开源项目推荐（47 个）

lemonadesaltbagel /crab

CRAB (Clean Rebuild of Anything's Behavior): rebuild software from behavior, not source code. An MCP-powered toolkit for black-box observation, clean-room reimplementation, and behavioral verification.

DefuzeX-AI /AgentBehaviorBench

AgentBehaviorBench (ABB): A benchmark for evaluating AI agents’ ability to test other agents and detect behavioral defects.

Basaltlabs-app /Gauntlet

Community-driven behavioral reliability benchmark for LLMs. 231 probes across 19 modules, deterministic scoring, perplexity correlation, layer sensitivity mapping, quant method capture, hardware-stratified community rankings. Every test contributes to the community dataset.

qualixar /agentassert-abc

Runtime behavioral contracts for AI agents: define YAML contracts, check adapter outputs, and raise explicit contract violations. Agent Behavioral Contracts (ABC).

Kartik8Dwivedi /Ditto

Execution-grounded semantic clone detection for JS/TS — finds functions that behave the same but are written differently, then runs them in a sandbox to prove which one is wrong.

datzle123 /RealDone

Prove web features work in a real browser—with persistence, source-of-truth, replay, and CI evidence.

GenesisClawbot /llm-drift

LLM drift detector — know within 5 min when GPT-4o, Claude, or Gemini silently changes behaviour. Open source, self-hostable.

deep-diver /ShapingBench

Executable behavioral contracts and hidden-evaluation harness for coding agents across nine software domains.

mohamed-khairy-5i /trigger-doctor

🩺 Behavioral testing for skill triggers — diagnose why a skill never fires (or fires too often), prescribe a fixed description, save a regression suite. One SKILL.md, 79 agents (Claude Code, Hermes, Codex, Gemini...)

Modelscape /modelscape

A toolkit for controlled behavioral evaluation of LLMs

sapniyogi /ui-behavior-coverage

Static analyzer for React tests that finds UI behaviors exercised but not explicitly verified. Testing Library, Jest/Vitest, and Material UI.

dhrumilbhut /behavtest

Behavioral regression testing for AI applications: repeat LLM, agent and RAG test cases, and tell real regressions from noise.

stef41 /modeldiffx

Model behavioral diffing - compare LLM outputs across versions, detect regressions.

hieunguyenak92 /rea-pro-max

Understand the mechanism. Rebuild the behavior. Prove the result. Independent reverse-engineering MCP, skill and CLI for Claude Code and Codex: artifact analysis, experiment planning and behavioral verification.

stef41 /modeldiff

Behavioral regression testing for LLMs — diff, drift, fingerprint. Zero deps.

shimo4228 /skill-comply

Agent Skill for Claude Code that measures whether coding agents actually follow a skill, rule or agent definition: it generates scenarios from supportive to competing prompts, runs an agent on each, and reports compliance rates with classified tool call timelines.

leonardocamacho1983 /ethogram

Open-source behavioral contract testing for AI agents. Define critical behavior, run the real agent, and catch regressions.

abdul-hamid-achik /cairntrace

Behavioral browser-spec layer for agent-in-session use. Specs declare intent+outcomes; agents execute + heal via agent-browser or Playwright. CLI + MCP server, agent-neutral.

senaayy /Computational-Cognitive-Lab

该项目在 GitHub 上没有填写描述

harman-04 /mockito-spies-and-verification-demo

Advanced Mockito usage featuring Spies, Mocks, and behavioral verification to test a shopping cart checkout flow.