open-source

NakshGuard is a self-hosted reverse proxy for OpenAI and Anthropic chat APIs that detects and blocks runaway AI agent loops before they waste tokens or API spend. It includes per-agent session tracking, configurable thresholds, shadow mode, and hot config reload.

A local-first, privacy-preserving usage meter for AI coding tools. It reads local transcripts to estimate token usage, cost, focus time, and achievements across tools like Claude Code, Codex, Kimi, and Qwen CLI.

Jo is a secure programming language designed to help catch prompt injection and other unsafe behavior at compile time by enforcing explicit capability boundaries. The repository shows the language’s security model, code examples, installation flow, and current project status.

Peers is a Python-based multi-agent coding substrate that coordinates Claude Code, Codex, and other CLIs to work together on coding, research, debugging, and repo audits with measurable convergence checks.

Open-source Claude Code plugin and skill for saving a compact handoff before hitting rate, usage, or context limits, then resuming in a new session without replaying the full transcript.

Open-source Python DDoS traffic monitor that prints live traffic stats to stdout and can output JSON for pipelines and tooling.

OpenClaw is a self-hosted personal AI assistant that runs on your own devices and connects to many messaging channels, with CLI-based setup and onboarding.

A survey of eight tools that sit between AI agents and the services they call, covering credential injection, interception, policy enforcement, and mocking/testing. The page explains what each tool is for, where it fits, and what to watch out for.

MagesticAI is an open-source, browser-based platform for managing AI coding tasks with multi-agent orchestration, task boards, terminal access, and code review. The README shows installation steps, Docker deployment, supported platforms, and a concrete local setup path.

Open-source shared memory for AI coding agents that records decisions, rejected alternatives, and contradictions across sessions.

Permission-first CLAUDE.md and agent stack for Claude Code and Google Antigravity, with 10 agents, 28 skills, and one-command installation.

A walkthrough for building and deploying an AI agent on Truuze with the AaaS dashboard, including setup, testing, connector configuration, and the platform’s escrow-based transaction flow.

An open-source Claude Code skill set for academic research, covering research, writing, review, revision, and finalization. It includes guided workflows, installation steps, and pipeline documentation for researchers using AI in a human-in-the-loop process.

An open-source Llama2 inference engine written in x86 real-mode assembly that fits in 1277 bytes and can boot directly from disk before any OS loads.

Open-source local-first terminal for working with terminal sessions and coding agents. It keeps retained PTY sessions alive on your machine, discovers sessions from several local agent tools, and exposes them through a browser-based dashboard.

Neuro is an AOT-compiled programming language for AI workloads, built on LLVM and positioned around native performance, tensor work, and future MLIR/Enzyme/GPU support.

Neoswarm is an open-source terminal workspace for controlling AI agents, with a plugin-based interface and typed TypeScript plugin API.

JarvisCore is a Python framework for building distributed multi-agent systems with peer-to-peer discovery, crash recovery, zero-trust credential handling, shared memory, and built-in observability.

A small Python CLI that hooks into Claude Code to gradually increase target-language usage during chat sessions, with monitoring, ramp policies, help modes, and a pause/kill switch.

Local-first coding agent workbench that runs agents in secretless microVMs, with support for multiple providers, parallel sessions, and shared repository workflows.

Terminal-native attention router for parallel Claude Code sessions. It shows session status in the terminal, highlights the one that needs help, and provides quick jump shortcuts, tmux support, and optional desktop notifications.

BOUND is a deterministic control harness for AI agents. It sits between agent execution and the next decision, using observable evidence to choose ACCEPT, RETRY, REPLAN, or ROLLBACK.

A small open-source C# tool for local inter-agent collaboration in coding workflows, with shared chat state, role memory, Markdown messages, and goal tracking inside a Git repo.

Open-source CLI for spinning up disposable local or remote sandboxes for AI agents. It uses a manifest and Dockerfile to build fresh boxes, then exposes a single shell tool through MCP so an agent can work inside the sandbox without touching the host.