Skip to main content

​
Devic AI Skills

Skills are reusable knowledge packages that provide AI coding assistants with detailed API documentation and usage patterns for the Devic platform. When installed, skills give your AI assistant deep understanding of Devic’s APIs, enabling it to help you build integrations more effectively.

​
Installation

Install Devic skills using the skills CLI: 
npx skills add devicai/skills
This command installs the complete Devic API skill package into your project, making it available to Claude Code and other compatible AI assistants.

​
What’s Included

The devicai/skills package includes comprehensive documentation for:

Agents API

Create, manage, and run autonomous AI agents with tool access and execution control.

Assistants API

Build conversational AI experiences with message processing and chat history.

Tool Servers API

Configure external tool integrations that agents and assistants can use.

​
How Skills Work

When you install a skill, it adds markdown documentation files to your project’s .claude/skills/ directory. These files are automatically loaded by compatible AI assistants, giving them context about the APIs and best practices.

​
Directory Structure

After installation: 
your-project/
├── .claude/
│   └── skills/
│       └── devic-api/
│           ├── SKILL.md          # Main overview and architecture
│           ├── agents.md         # Agents API documentation
│           ├── assistants.md     # Assistants API documentation
│           └── tool-servers.md   # Tool Servers API documentation

​
Quick Start

​
1. Install the skill

npx skills add devicai/skills

​
2. Get your API key

Generate an API key from the Devic dashboard. Keys follow the pattern: 
devic-{random_string}

​
3. Set up authentication

Include your API key in all requests: 
Authorization: Bearer devic-your-api-key

​
4. Start building

Your AI assistant now understands the Devic API. Try asking it to:
  • Create an agent with specific tools
  • Set up a tool server for your external API
  • Send messages to an assistant
  • Monitor agent execution threads

​
Architecture Overview

Understanding how Devic entities relate to each other: 
┌─────────────────────────────────────────────────────────────────┐
│                         Tool Server                             │
│  (External API integration with tool definitions)               │
└─────────────────────────────────────────────────────────────────┘
                              │
                              │ referenced by
                              ▼
┌─────────────────────────────────────────────────────────────────┐
│                       Tools Group                               │
│  (Logical grouping of tools - built-in or from tool server)     │
└─────────────────────────────────────────────────────────────────┘
                              │
                              │ availableToolsGroupsUids
                              ▼
┌─────────────────────────────────────────────────────────────────┐
│              Assistant Specialization                           │
│  (Configuration: presets, tools, model, provider, etc.)         │
└─────────────────────────────────────────────────────────────────┘
                    /                   \
                   /                     \
                  ▼                       ▼
┌──────────────────────┐     ┌──────────────────────────────────┐
│     Assistant        │     │            Agent                  │
│  (Chat interface)    │     │  (Autonomous execution threads)   │
└──────────────────────┘     └──────────────────────────────────┘
         │                              │
         ▼                              ▼
┌──────────────────────┐     ┌──────────────────────────────────┐
│   Chat Histories     │     │          Threads                  │
│  (Conversations)     │     │  (Execution sessions with tasks)  │
└──────────────────────┘     └──────────────────────────────────┘

​
Base URL

All API requests are made to: 
https://api.devic.ai
Example endpoint: 
GET https://api.devic.ai/api/v1/agents

​
Response Format

​
Success Response

{
  "success": true,
  "data": { ... },
  "timestamp": "2024-01-15T10:30:00.000Z"
}

​
Error Response

{
  "success": false,
  "error": {
    "message": "Error description",
    "code": "ERROR_CODE"
  },
  "timestamp": "2024-01-15T10:30:00.000Z"
}

​
Next Steps

Agents API

Learn how to create and manage autonomous AI agents.

Assistants API

Build conversational AI experiences.