28 October 2026

A collection of design skills for AI coding tools

A collection of standalone skill folders for Claude Code, Cursor, Codex, Windsurf, and other tools that read markdown rule files. For now there is only one skill: landing-page-design.

• landing-page-design builds a landing page from scratch: initial questions, page structure, conversion copy, and a visual system (typefaces, spacing, radius, motion).
• For Cursor, SKILL.md is copied to .cursor/rules. For Codex, to AGENTS.md, and for Windsurf, to .windsurfrules.
• Design values such as fonts, colors, spacing, and motion are in a single section near the end of each skill file.

A design skill that avoids the AI-generated look

A skill from Together AI for Claude Code, Codex, Cursor, and OpenCode. The build output is linted in a real browser against signs of a machine-made look.

• Offers three themes from a catalog of 50, then assembles a page around the visitor's goal.
• There are three gate levels: Floor must not be bypassed (contrast, focus, reduced motion, broken mobile layout), Reflex can be waived with a reason, and Finish does not block a release.
• A reviewer with fresh context gives the final verdict.

Desktop workspace for persistent AI agents

A local-first desktop app. Each agent has its own workspace and conversation. Supports Codex, Claude Code, Grok CLI, and others through ACP.

• A FIFO message queue with pause, resume, and cancel, plus file transfer and agent-to-agent messaging.
• An embedded browser that agents can open and control.
• Still a development preview: agents run with danger-full-access and approvalPolicy: never, with no per-action confirmation.

Software development methodology for coding agents

A set of composable skills, plus initial instructions so the agent uses them. Skills trigger automatically, with no special commands.

• The agent first asks about the goal, then shows the spec in short chunks for approval.
• After approval, the agent drafts an implementation plan and runs subagent-driven-development with TDD red/green, YAGNI, and DRY.
• According to the README, the agent often works autonomously for a few hours without deviating from the plan.

Design guidance for AI coding agents

A single skill with 24 commands and 62 detector rules for spotting signs of AI-generated frontend design. A CLI and a browser extension are available.

• /impeccable init inspects the project and writes PRODUCT.md: audience, goals, context, constraints, and voice.
• Commands include polish, audit, critique, distill, animate, bolder, and quieter.
• The detector rules run without an LLM and without an API key.

Ready-to-use open-source AI agent

Connects a computer and a phone in one session, using the models and subscriptions you already have. Can use Claude Code, Codex, and Pi.

• A session can be continued from another device, including access to its files.
• A built-in remote desktop on the phone to open apps and type on the connected computer.
• Can lead a team of agents: planning, parallel implementation, and review by different agents.

Social media scheduling and analytics platform

Manages scheduling, automation, and analytics for content. Available as managed Postiz Cloud or a self-hosted open-source edition.

• Team members can collaborate, comment, and schedule posts.
• Has a Public API, a NodeJS SDK, a custom N8N node, and Make.com integration. Zapier is also mentioned.
• Stack: Pnpm monorepo, NextJS, NestJS, Prisma (PostgreSQL by default), and Temporal.

Coordinated engineering system for coding agents

Adds a plan, test, implement, review, verify, remember, improve flow to agents. MIT license, works best with Claude Code.

• Includes 71 agents, 311 skills, and 97 legacy command shims, plus hooks, rules, memory, and the AgentShield security scanner.
• Has native plugin paths for Codex and Qoder, and limited adapters for Cursor, OpenCode, Gemini, Zed, and others.
• The README asks that you install only from official channels: the GitHub repo, the npm package ecc-universal, and the ecc.tools site.

Personal AI tutor for long-term learning

A tutoring app built with Python and Next.js, with a web interface and a CLI. Latest release in the README: v1.6.15, 2026.10.19.

• v1.6.15 adds reviewed learning path changes, grouped assessments, bounded tutoring memory, and native Gemini voice.
• Earlier releases added Reading, practice from a question bank, knowledge bases, and a study journal.
• The README describes the CLI as an agent-native interface, and EduHub as a community for skills.

Agentic IDE built around the pi coding agent

From JetBrains. An IDE interface for the pi agent, with isolated workspaces and project knowledge through living specs and skills.

• Each workspace is a git worktree with its own branch, so agents work in parallel without touching your checkout.
• Linked SPEC.md files next to the code are read and maintained by the agent through the spec_* tools.
• Includes a Monaco editor, a file tree, a terminal, and review comments on diffs that can be sent to the chat.