Skills.sh catalogs reusable skills and simplifies installation for supported coding assistants. Its directory lets developers review use cases before adding capabilities to their workflow.
Tech watch
← All linksAI Agents
Page 3 of 3
NullClaw aims to provide compact infrastructure for an autonomous assistant, written in Zig. Its repository makes the project and implementation choices easy to explore.
This skill gives an agent design guidance and interface choices for multiple platforms. The repository includes style, typography, and color-pairing recommendations to shape an implementation.
Rowboat presents itself as a collaborative assistant with memory and tool integrations. Its repository makes the architecture and integrations available for inspection beyond the product overview.
This article shows how to expose a Storybook-documented React library through a local MCP server. Component manifests let an agent consult design-system rules before generating code.
React Grab lets you select a visible element in an app and pass its context to a coding agent. It helps locate the corresponding component in the codebase.
Claude-Mem captures session activity, compresses it, and brings relevant details into later sessions. Its repository describes persistent memory for several coding agents.
Trigger.dev runs long-lived tasks and agents with retries and concurrency controls. Its TypeScript SDK keeps those workflows close to application code.
Vercel organizes guidance around major sources of slowness: waterfalls, bundles and unnecessary renders. Prioritization helps agents address costly problems before micro-optimizations.
Vibium exposes navigation, forms, clicks and screenshots through CLI, MCP or client libraries. It uses WebDriver BiDi so agents can verify changes in a real interface.
Firecrawl offers search, scraping and crawling with Markdown, JSON and screenshot outputs. Its API provides an extraction layer for workflows and agents using web content.
Firecrawl
Mastra combines tools, workflows, traces and evaluations for building TypeScript agents. The framework supports observing execution and improving behavior through experiments.
This practical course guides the creation and deployment of Mastra agents. Tools, memory and MCP are introduced through a guided workflow inside the code editor.
This library derives LLM tools from React hooks. Human and agent interfaces can share the same state, with a developer-defined schema controlling state changes.
The repository introduces the Codex CLI coding agent and installation procedures. It provides an entry point to a code-assistance workflow running locally on the machine.
This server exposes the Node.js debugger to assistants such as Cursor and Claude Code. It adds runtime information to the context available for diagnosing problems.
This extension combines discovery, installation and management of MCP servers and skills. It makes those resources available to coding agents used alongside the editor.
Midscene uses a visual agent to perform actions and verify interface results. Tests can be described in YAML or integrated into an existing framework through its APIs.
PR-Agent provides a community-maintained agent for reviewing pull requests. The repository documents setup and distinguishes this open-source project from Qodo’s commercial offering.