drift-ci core engine: types, providers, evaluators, storage, runner

Architecture policy as one manifest of your repository, language-agnostic: the manifest schema, the evaluators for imports, exports, members, surface, structure and the import graph, the ports a language pack implements, and the loader that turns a manife

Habenula governance kernel — the pure permission and spending evaluators. Leaf package: no runtime dependencies.

CLI for interacting with the Google Genkit AI framework

Various evaluators for the smartshop platform.

Genkit AI framework plugin for RAG evaluation.

Evaluation framework for AI Copilot applications - versioned datasets, execution records derived from recorded diagnostics, deterministic evaluators (tools, permission compliance, RAG, citations, memory, agents, workflows, UI, latency, tokens, cost), base

General-purpose GPU data and computation for luma.gl

Version-controlled, replayable eval cases for AI agents: ingest traces, store YAML cases, score with pluggable evaluators, fail CI on regression.

Much like tests in traditional software, evals are an important part of bringing LLM applications to production. The goal of this package is to help provide a starting point for you to write evals for your LLM applications, from which you can write more c

Hadron's authorization core — the pure, isomorphic access-control evaluators used by hadron-server and every Hadron client. The server is always the enforcement authority; clients use this to predict its decisions.

Generic class to process and serialise universal Turing machines and evaluators

Lightweight deterministic evaluators for AI agents. Binary pass/fail checks, zero dependencies, no LLM cost.

Visual Oracle for Agentic Evaluators - Semantic UI testing using LLMs.

This is a collection of rules, rule-evaluators, and tests for semantic validation of Versa receipts. Written in Rust, it uses [napi-rs](https://napi.rs/) to compile to native modules for use in NodeJS environments. It can also be used in Rust backends, an

Testing utilities for Agent OS — CSV test runner, evaluators, and test suite tools

TypeScript definitions for multisort

Plain authored journey graphs, timed evaluators, cues, and retiming helpers for Found in Space

A library for running evaluations for AI use cases

Daily Mission & Check-in event module for Echo games. Players complete missions to earn Sound Waves that fill stage chests, progressing toward a Grand Prize. Supports duration and weekly-schedule modes, auto/manual start, pluggable mission-type evaluators