AI Security using FOSS tools with Macbook Pro --> Laya + NOVA prompt protection with Gateway API implementation on Kubernetes

AI Agent 环境安全审计 CLI：供应链投毒 / 密钥暴露 / 端口暴露 / MCP server 一键体检（零依赖、只读、Windows 优先）

A corpus of Lean 4 theorem statements harvested from open formalization libraries and CompeteMath's own proven problems, staged for the Leak prover to attempt.

Lightweight Outlook Autodiscover, Thunderbird Autoconfig and Apple Mail profile service for self-hosted mail servers.

Local relay that lets one Claude Code session run each subagent tier on a different subscription, passing your own OAuth through untouched.

Python security utility for detecting and redacting likely secrets at runtime, in source files, Git commits, and CI.

OpenVINO Physical AI provides the deployment-side components for running trained policies on real hardware. It handles camera capture, robot control, and...

The passmcp attestation format for MCP servers, its JSON Schemas and offline verifier, the security graph model, and an agentgateway processor that enforces it.

The open source runtime proxy that instantly transforms enterprise API specs into secure, fine-tuned, and consistent MCP servers for LLMs and AI agents.

Credentials and APIs for you and your coding agents: AWS roles with MFA, OAuth and API keys, an OpenAPI explorer, and data and database access, without...

Standalone OpenID Shared Signals Framework transmitter for self-hosted Authentik deployments.

Centralized live status board for GCP, AWS, Steam, CS2 Europe, Epic/Fortnite, Spotify, Apple, Android, Grok, ChatGPT, and Claude.

A blazing fast, type-safe library for working with JSON in Swift

Personal Infrastructure as Code. Declarative, reproducible, and versioned NixOS workstation managing system, Home Manager, secrets, storage, services, etc.

Pure-Python reader/writer for VBA macros embedded in Excel / PowerPoint / Word / Access files. Zero dependencies, deterministic round-trip, safe on protected...

Unit testing and build system for C projects

Watch, decide, enforce and record the tool calls your AI agents make: an MCP gateway with signed policies, human approvals, pause and an evidence trail. Open...