Articles, releases and code from Hacker News, Reddit, GitHub and the people building RPA, workflow automation and AI agents — plus what the community pushed to the top today.
AIShipback released an SFTP deployment tool and MCP server that lets AI agents safely deploy to shared hosting with automated backups and one-command rollbacks.
AIA solo founder detailed deploying 17 specialized AI agents with separated contexts to handle publishing, trading diagnostics, and ERP administration across a distributed system architecture.
AIAn agent workflow caught false success logs by validating actual email delivery, reminding builders to verify real end-state outcomes rather than trusting returned API status codes.
AIApple research shows minimal coding agents match complex multi-agent harnesses, allowing developers to simplify agent architecture and rely on frontier models rather than elaborate orchestration layers.
AIAstrax released an open-source MCP server that lets AI models automate desktop GUI and terminal tasks across operating systems using local state management to prevent context bloat.
AISakana AI released Agentic Scientist, an open-source orchestrator that directs coding agents to autonomously run end-to-end scientific research workflows from idea generation through paper writing.
AIRadxa's Linkr is a compact Wi-Fi KVM device that provides hardware-level keyboard, mouse, and video control, allowing AI agents to automate physical computers without native network access.
AIAIOps Cockpit released an open-source AI SRE tool that investigates infrastructure incidents and proposes remediation diffs requiring two-person human approval before execution.
AITrustbut is a local CLI tool that verifies AI coding agent outputs by checking whether newly generated tests actually fail without the agent's code changes.