> ## Documentation Index
> Fetch the complete documentation index at: https://docs.stagewise.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Skills & plugins — Extend stagewise agents

> Extend AI coding agents in the stagewise agentic IDE with custom skills (SKILL.md) and built-in plugins. Integrate Figma, GitHub, Supabase, PostHog, Vercel, and Remotion.

Skills and plugins extend stagewise beyond its default capabilities. Skills teach the agent about your project. Plugins add new tools, integrations, and interactive features.

## Skills

A skill is a set of instructions the agent follows when working on specific tasks. You create skills as `SKILL.md` files in your workspace.

### How skills resolve

When the agent picks up a task, it checks whether any skills match the trigger, then resolves them in priority order:

```mermaid theme={null}
graph TD
    A[Agent receives task] --> B{Skill trigger<br/>matches?}
    B -->|Yes| C{Where is the skill?}
    B -->|No| H[Use default agent behavior]
    C --> D[.stagewise/skills/<br/>Highest priority]
    C --> E[.agents/skills/<br/>Shared across tools]
    C --> F[Built-in plugin skills]
    D --> I[Agent follows skill instructions]
    E --> I
    F --> I
```

Higher-priority versions shadow lower ones when the same skill exists in multiple locations.

### What skills do

Skills can teach the agent:

* **Project conventions** — Naming patterns, file structure, preferred libraries
* **Domain knowledge** — How your framework or system works
* **Workflows** — Multi-step processes the agent should follow
* **Constraints** — Rules the agent must respect

### Skill locations and priority

| Location             | Priority | Scope                  |
| -------------------- | -------- | ---------------------- |
| `.stagewise/skills/` | Highest  | stagewise-only         |
| `.agents/skills/`    | Lower    | Shared across AI tools |

### Skill structure

```
.stagewise/skills/my-skill/
├── SKILL.md          # Instructions (required)
├── references/       # Files the agent can read
└── assets/           # Supporting files
```

A SKILL.md needs a clear trigger description:

```markdown theme={null}
---
name: "tailwind-patterns"
description: "Use when creating or modifying React components with Tailwind CSS."
---

# Instructions

Always use design tokens from the theme, never hardcoded colors...
```

### Slash commands

Invoke skills explicitly with slash commands. Type `/` in the chat to see available commands from your skills and plugins.

## Plugins

Plugins are pre-built extensions that can include skills, credential storage, and interactive mini-apps.

### Built-in plugins

| Plugin       | What it provides                                     |
| ------------ | ---------------------------------------------------- |
| **Figma**    | Access Figma designs, real-time selection monitoring |
| **GitHub**   | Repository access, issues, pull requests, actions    |
| **Supabase** | Database queries, edge functions, migrations         |
| **PostHog**  | Analytics queries, feature flags, experiments        |
| **Vercel**   | Deployments, logs, environment variables             |
| **Remotion** | Video creation and editing in React                  |

### Plugin credentials

Some plugins need API keys or tokens. Credentials are encrypted on your machine and never exposed in plaintext. Configure them in **Settings → Agent → Plugins**.

### Plugin apps

Plugins can ship interactive mini-apps that render inside the chat sidebar. The agent opens these apps and exchanges data with them — for example, a Figma plugin showing a design preview panel.

## Skills vs plugins

|                | Skills                | Plugins                   |
| -------------- | --------------------- | ------------------------- |
| **Created by** | You, for your project | stagewise                 |
| **Scope**      | Per workspace         | Global (all workspaces)   |
| **Includes**   | Instructions only     | Skills, credentials, apps |
| **Stored in**  | Your workspace        | stagewise installation    |

## What's next

<CardGroup cols={2}>
  <Card icon="lightbulb" href="/core-concepts/agent-context" title="Agent context">
    Understand how skills fit into the agent's context system.
  </Card>

  <Card icon="settings" href="/reference/skills-and-context" title="Skills & context reference">
    Configure skills and context files per workspace.
  </Card>

  <Card icon="puzzle-piece" href="/reference/plugins" title="Plugins reference">
    Manage installed plugins and configure credentials.
  </Card>
</CardGroup>