Desktop Commander MCP, a tool that gives AI agents structured control over a computer via API calls rather than vision-based interaction, has gained 6.5K GitHub stars with rapid growth. The tool enabl…
An engineer implemented a zero-dependency MCP (Model Context Protocol) server in pure Python using only the standard library, avoiding the 47-package dependency tree of the official SDK. The implement…
Selvedge, a new open-source tool, provides long-term memory for AI-coded codebases by capturing the reasoning behind code changes made by AI agents like Claude Code, Cursor, and Copilot. It stores the…
A developer built Sparsi, a framework that shifts complex logic out of ReAct agent prompts into deterministic DAG-based macro-tools. In a benchmark on customer support ticket triage, Sparsi achieved 1…
A developer using Claude Code daily identified three costly habits: Claude claiming completion without verifying code, excessive context token usage from auto-loaded files, and lack of cross-session m…
A developer tested whether small local language models can drive MCP (Model Context Protocol) tools using Goose's built-in local inference on an M3 24GB laptop. A 4B model successfully executed tool c…
The developer team behind Chron, an MCP server that records AI coding sessions to a local SQLite database, has shipped a SOC 2 review layer. The tool scans session history for findings against SOC 2 T…
Didit shipped Copilot, an AI agent embedded in its Business Console that lets users manage identity verification workflows via natural language. The agent operates the console in real time, using thre…
A developer open-sourced FableCut, a browser-based video editor designed for AI agents to operate via a JSON project file. The entire timeline lives in a single document, allowing any tool that writes…
A developer built agentproto, a daemon and CLI that unifies the lifecycle of multiple AI coding agents—Claude Code, Codex, Hermes, opencode, and Mastra—under a single interface with durable supervisio…
An autonomous agent has built and maintains a leaderboard ranking 9,868 MCP servers by GitHub stars, npm downloads, and maintenance activity, updated every other day without human intervention. The ra…
A developer built a browser-based agent that reverse-engineers web apps' own APIs to automatically generate reusable tools for AI assistants, enabling deep integration without modifying source code. T…
Hamza built Satori, an open-source, read-only codebase map for MCP-compatible AI coding agents. It provides a structured investigation path to help agents efficiently navigate symbols, ownership bound…
Agentics introduces design principles for building command-line interfaces (CLIs) optimized for AI agents as primary users, emphasizing no colored output, no interactivity, recursive help messages, an…
The Linux Foundation's Agentic AI Momentum Report tracks 116 open source agentic AI projects across five layers, finding 2.6x growth in unpatched CVEs over six months and organizational concentration …
Agentic FC, a football management simulation designed for AI agents to play via the Model Context Protocol (MCP), has been released as an open-source Go project. The game features a terminal-based spe…
Google's WebMCP protocol, announced at Google I/O 2026, is a browser-scoped tool protocol for agents operating inside a Chrome tab, not a replacement for existing MCP servers. A developer running thre…
Oomol Lab released OpenConnector, an open-source project that provides a runtime boundary for SaaS access in AI agents, addressing issues like user account management, scope enforcement, and schema co…
Focused Labs argues that AI agent security must be enforced at the tool call, not at connection setup. The HCP paper defines eight runtime invariants for MCP-style systems, and its benchmark shows tha…
A developer ran 70 MCP servers in a sandboxed environment using Docker and strace to log runtime behavior, finding that 67 were clean at startup, 3 made outbound HTTPS connections consistent with thei…