All categories
Generative Code

Screenery is a tool for managing product screenshots that fits into your continuous integration (CI) workflow. It automatically uploads screenshots with every build, assigning each one a stable URL that ends in @latest.png and corresponds to your default branch. When you merge UI changes, all pages using that URL will display the updated screenshot without needing manual updates. Screenery also offers visual regression testing, which stops the CI process if unexpected changes are detected before they affect your documentation. It comes with an MCP server and a GitHub Action designed for use with coding assistants like Claude Code, Codex, and Cursor.
DocQuery integrates Excel workbooks, Google Sheets, or Airtable databases with PDFs, contracts, and invoices. It employs artificial intelligence to extract information and organize it into relational tables with foreign keys. The data obtained is accessible via REST, GraphQL, and a Model Context Protocol server, which is compatible with Claude and Cursor. Updates in the source spreadsheets are automatically synchronized back to the API, ensuring data is always up to date. Users can filter, paginate, and navigate through related rows without needing to write SQL or handle migrations.
The Forge by Eternitech is a web effect generator powered by AI. It transforms instructions written in plain English into high-performance vanilla JavaScript injection tags within a secure environment. When users describe a visual effect, the tool swiftly compiles it and deploys it to the Cloudflare R2 edge CDN as a single, universal script tag. This tool is compatible with widely used platforms such as Webflow, Framer, Shopify, WordPress, and React. Before deployment, each generated script undergoes automatic testing to check for memory leaks, CPU issues, and DOM pollution.
Kai is an AI-based code review tool that examines GitHub pull requests by comparing them to a semantic graph of your complete codebase. It detects bugs and architectural problems that go beyond what is visible in the diff. Kai maps symbols, callers, dependencies, and test relationships to understand how changes affect the code, not just which lines have been altered. Before presenting any issues, Kai verifies each finding, ensuring that only confirmed problems appear in its inline PR comments. Additionally, Kai offers a command-line interface (CLI), a cloud agent for automated corrections, and an MCP server that provides AI coding assistants with access to call graphs and impact analysis. The pricing options range from a free tier to paid plans at $15, $50, and $100 per month.
Levanto Sage2 is an API aimed at developers. It enhances applications that deal with natural language input by adding features for extracting information and making streaming decisions. Sage2 transforms unstructured messages from customers into organized data types like text, numbers, monetary values, dates, and durations. Each result includes source links and confidence levels. Applications can use streaming decision sessions to evaluate up to 32 questions based on the ongoing conversation context. Answers are updated in real time as new input comes in. Sage2 is built to work with language models, providing fields and decisions while the application code manages validation and action execution. Additionally, Sage2 supports reasoning, web grounding, and vision capabilities, which are currently in beta.
Totem is a collaborative workspace platform that integrates AI agents from various tools and accounts into a unified, coordinated network. Users can set up Missions to define objectives, manage task transfers, and share information among agents such as Claude Code, Codex, Cursor, and others. Each agent connects through MCP or API, and owners can finely control permissions, secure requests, and the sharing of files or context. Totem offers a centralized view of agent activities, messages, and progress, enabling teams to monitor work in real time.