find the tool
for the thing you
are building right now

MCode Stack is an open-source collection of workflows, skills, and checklists used by the MiniMax Code team to maintain high-quality AI-generated code. It provides practical, daily-use engineering habits that can be installed directly into the MiniMax Code agent to improve task management, context handling, and code review processes.

Engine is a high-performance, local AI infrastructure designed specifically for Apple Silicon. It enables developers to build applications that can read, see, hear, speak, and create entirely on personal hardware without requiring internet connectivity or cloud-based API calls.

OpenDuet is an open-source, self-hosted template for proactive AI agents designed to manage tasks autonomously while keeping users informed. It features a 'computer of its own' to execute workflows in the background, provides daily morning briefs, and implements a human-in-the-loop approval system for sensitive actions.

Supabase Compute is a managed service that allows developers to run AI agents, background jobs, and web services directly within their Supabase project. By providing a full Linux environment, it enables the execution of long-running tasks and services in the same network as the project's Postgres database.

Agent SDK is a development framework that allows users to define, test, evaluate, and deploy AI agents as code. It leverages the same runtime and harness that power Cursor's internal tools, enabling developers to build custom agents using a combination of Markdown instructions and TypeScript.

Papers with Code is an open-source platform that tracks trending AI research by connecting academic papers with their official code implementations, datasets, and benchmark leaderboards. It serves as a community-driven successor to the original Meta-operated platform, providing a comprehensive catalog of machine learning artifacts and research tools.

Devin Security 2.0 is an autonomous security platform that integrates frontier-model scanning with a living threat model to identify and prioritize vulnerabilities. It features autonomous remediation capabilities that generate, test, and submit pull requests directly to your codebase to resolve security findings.

Lifeguard is a static analysis tool developed by Meta designed to detect code patterns that are incompatible with Python's Lazy Imports. It helps developers identify and resolve issues to ease the adoption of deferred import semantics in preparation for Python 3.15.

FrontierMemory is an open-source benchmark designed to evaluate how well AI agents learn from past experiences rather than simply recalling facts. It measures the effectiveness of memory generation by testing whether an agent's learned memory folder improves performance on 15 downstream coding and knowledge-work tasks.

The Single agent loop kit is a reference implementation designed to help developers transition from complex multi-agent architectures to a more efficient single-agent loop. It provides a structured, data-driven workflow to audit, measure, and fold subagents into skills and tools within a main agent.

Agensh is a decentralized, self-organized multi-agent harness designed to scale up to 1,024 agents without a central orchestrator. It enables agents to autonomously discover and claim sub-tasks, share findings, verify results, and merge progress asynchronously through a shared workspace and messaging interface.

Vercel CLI Spend Management allows teams to monitor and control their cloud infrastructure costs directly from the command line. Users can view current spending, set custom budget thresholds, and configure automated actions like pausing production deployments when limits are reached.

Reason Agent 1.1 is an autonomous software engineering agent designed to perform coding tasks by interacting with repositories, running tests, and opening pull requests. It optimizes performance and cost by utilizing a specialized harness that enables parallel tool execution and efficient token usage.

AWS Continuum for code vulnerabilities is an AI-powered security service that automatically discovers, validates, and prioritizes code vulnerabilities within your environment. It generates actionable remediation fixes, such as patches or pull requests, and provides proof-of-concept evidence to help teams resolve security issues at machine speed.

Microsoft.UI.Reactor is an experimental, declarative C# framework for building WinUI 3 desktop applications. It provides a React-style programming model with hooks-based state management, allowing developers to build native Windows UIs without XAML, data binding, or ViewModels.

The Anthropic Sandbox Runtime is a lightweight, open-source tool designed to enforce filesystem and network restrictions on arbitrary processes at the OS level. Developed primarily for Claude Code, it provides a secure environment for running AI agents, local MCP servers, and shell commands without the overhead of traditional containers.

rl-env-use is an open-source tool that converts a one-line skill description into a verified reinforcement learning environment. It leverages Browser Use agents to mine real-world cases from the web, creating datasets and graders that help train models and optimize prompts.

RL360 is an open-source reinforcement learning infrastructure designed for the post-training of language models, specifically powering the K2 Horizon model family. It enables training models on complex tasks involving coding, tool use, and interaction with sandboxed environments through a distributed Ray-based architecture.

The Gemini API CLI is an experimental command-line tool designed to provide coding agents and shell scripts with direct access to Gemini models and managed agents. It enables developers to perform tasks like media generation, file analysis, and background agent orchestration directly from the terminal.

The Supabase .NET SDK is a comprehensive C# client library that provides seamless integration with Supabase services including Auth, Database, Storage, Realtime, and Edge Functions. Version 8.0.0 introduces significant architectural improvements, including a migration to System.Text.Json and updated framework targeting for modern .NET development.