Agent Skills

SKILL.md packages that extend Claude Code, Cursor, Copilot, and other AI agents.

Trust Score Usefulness Recommended Stars Latest

Filters

Code Mode for MCP Servers

code-mode-skill

Add a sandboxed code mode tool to an MCP server so LLMs run small processing scripts against large API responses and only the script output enters the model con

mcpsandboxcode-execution

6 triggers

CE Work (Beta)

Compound Engineering Plugin

Execute development work from a plan or prompt with a structured, test-first workflow and optional Codex delegation for automated implementation.

engineeringworkflowautomation

21,527

5 triggers

Figma Pilot

figma-pilot

Reference and operator guide for figma-pilot: how to use figma_execute and related APIs to programmatically create and modify Figma designs.

figmadesignmcp

5 triggers

E2B — Sandboxed Code Execution

TerminalSkills Library

Documentation and best-practices for E2B sandboxes: run AI-generated code, install packages, manage files, and return artifacts safely.

sandboxcode-executionagent-tools

105

5 triggers

Codex CLI Delegate

codex

Delegate focused coding tasks to the Codex CLI via a wrapper script; use only when the user explicitly asks Codex to act.

clicode-executiondeveloper-tools

5 triggers

Codex Agent Runtime for OpenClaw

codex-agent

Managed runtime skill for driving OpenAI Codex CLI from within OpenClaw. Supports interactive tmux sessions, one-shot exec tasks, session status queries, explic

codexopenclawcoding-agent

160

8 triggers