Grabbit launches agent-first screenshot API with MCP support and one-line CLI
Framejet is a screenshot API that takes a URL and returns a PNG/JPEG, removing cookie banners, sticky headers, and live-chat widgets before capturing. It exposes a remote MCP server so AI agents like Claude and Cursor can call it as a tool. It targets developers automating screenshots who otherwise get pages obscured by consent walls and chat bubbles.
Framejet is a screenshot API that takes a URL and returns a PNG/JPEG, removing cookie banners, sticky headers, and live-chat widgets before capturing. It exposes a remote MCP server so AI agents like Claude and Cursor can call it as a tool. It targets developers automating screenshots who otherwise get pages obscured by consent walls and chat bubbles.
Chaser is a new macOS utility that captures more than screenshots by using accessibility features to provide AI agents with UI context like layout, button states, and object mappings. It is 100% local with no LLM, inspired by Codex appshots, and includes history and annotation features.
Chaser is a new macOS utility that captures more than screenshots by using accessibility features to provide AI agents with UI context like layout, button states, and object mappings. It is 100% local with no LLM, inspired by Codex appshots, and includes history and annotation features.
Getdesign is an agent that takes a screenshot of any website and uses an LLM to generate a design.md file describing its design. It helps developers and designers capture design specs from existing sites.
Getdesign is an agent that takes a screenshot of any website and uses an LLM to generate a design.md file describing its design. It helps developers and designers capture design specs from existing sites.
A desktop screenshot tool: user marks an area via a shortcut and overlay, and it saves the highlighted crop plus full-screen context image and a markdown file with element info like browser URL. Built to streamline pasting annotated screenshots into LLM agent chats instead of manually explaining each image. Built by a developer for their own workflow with coding agents.
A desktop screenshot tool: user marks an area via a shortcut and overlay, and it saves the highlighted crop plus full-screen context image and a markdown file with element info like browser URL. Built to streamline pasting annotated screenshots into LLM agent chats instead of manually explaining each image. Built by a developer for their own workflow with coding agents.
A developer released a simple open-source utility called 'look' that allows Claude Code to see the user's screen displays. The tool, available on GitHub, lets users instruct Claude to 'take a look' and view all active displays, eliminating the need to paste screenshots manually.
A developer released a simple open-source utility called 'look' that allows Claude Code to see the user's screen displays. The tool, available on GitHub, lets users instruct Claude to 'take a look' and view all active displays, eliminating the need to paste screenshots manually.