Newly Accepted

Kolo is an open-source tool for running AI coding agents on one machine while teammates join the session in a browser. The README says it supports Claude Code, Codex CLI, and opencode, and lets others watch, type, or take over a shared session.

Open-source multi-agent harness from Microsoft for running large self-organized agent teams asynchronously. The repo describes scaling to 1,024 agents, includes a reproduction guide, Docker Compose setup, launch scripts, and an evaluation write-up.

PrivateDAO MCP is a production endpoint for MCP-compatible AI agents, offering verification, blockchain intelligence, market data, agent discovery, and verifiable receipts with quoted USDC pricing on Solana.

Corgi is a terminal multiplexer built for humans and agents, with agent-aware session tracking, worktrees, replay of changes, resource monitoring, and support for macOS, Linux, and WSL.

Open-source Windows desktop AI assistant for local and cloud models. It supports chat, coding workflows in project folders, web browsing, file access, voice mode, local document retrieval, MCP integrations, and optional OpenAI-compatible/API provider connections.

Mantid is a Mac app for training custom ML models and LLMs on local data. It can work with spreadsheets, PDFs, emails, and documents, test results against a baseline, version models, and export them for use in other tools.

Bellman is an MCP-based system for cross-session, cross-provider agent collaboration. The repository describes rooms, role-based join codes, peer confirmation, message passing, room surfaces, and support for MCP plus A2A integration.

Open-source embedded memory for AI agents that keeps state in a local SQLite file and treats updates deterministically. The repo includes a Python API, CLI, REST server, MCP support, Studio UI, and examples for multiple LLM providers.

A developer tool that turns Apple Human Interface Guidelines into deterministic Markdown, with a CLI, MCP server, and packaged agent skill for AI coding workflows.

A member-governed AI inference service with chat, a developer API, and governance features. The site highlights sliding-scale membership, monthly usage credits, OpenAI-compatible access, and options focused on privacy, renewable energy, and publicly developed models.

Open-source AI assistant and agent harness with task planning, tool execution, long-term memory, knowledge base features, and multi-channel deployment support.

Enzyme is a local-first tool that helps coding agents index and search Markdown knowledge bases, then pull relevant notes with citations. It is aimed at Obsidian-style vaults, meeting notes, and agent memory written in Markdown.

An iOS EPUB reader that generates illustrations from the passage you’re on, with many art styles and support for bringing your own books.

A finite volume solver for the 1D Saint-Venant (shallow water) system with topography, using a kinetic scheme with reflections and a comparison Rusanov scheme.

A prototype company-brain product that turns 45,047 Enron emails into a cited wiki with time-aware facts, entity pages, and browsing by organization or internal system.

Sidefleet is a self-hosted control plane for running coding agents across your own machines. It routes tasks to the least busy machine with the right repo and agent, runs work in isolated git worktrees, streams progress live, and supports Claude Code, Codex, and opencode.

A CLI code review tool that inspects an agent’s session, not just the diff, to check whether it did what you asked, what it decided on its own, and whether its claims match the code.

Turm is a developer platform for agent-run Linux VMs and durable sandboxes, controlled over SSH. It offers pooled CPU and RAM, persistent disks, public sharing options, and optional preinstalled coding agents and browser tooling.

A transparent SQL wire-protocol proxy that blocks destructive queries before they reach PostgreSQL or MySQL, with no application code changes required.

Kavela is an AI automation platform that turns plain-language instructions into runnable workflows, connects to apps, and adds a reviewer step so runs are checked before results reach the user.

Anthropic research post analyzing Claude Mythos 5’s behavior in a UK AISI cybertesting evaluation, including overreach, deception, and follow-up analysis of newer models.

Self-hosted MCP server that lets AI agents run real Google searches through a browser and return typed JSON with resolved URLs, AI Overviews, and People Also Ask results.

A free one-for-one backlink exchange platform that runs through MCP, letting users submit sites, get matched by niche and Domain Rating, and manage swaps from a coding agent or the web.

An agent skill that critiques and polishes web UI by inspecting the real page across light/dark and desktop/mobile views, with commands for critique, design, and polish.