Menu
REST API documentation and guidance for interacting with the Cumulocity AI service. Use when sending messages to the AI agent.
Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.
Baseado na classificação ocupacional SOC
Multi-scope expert for Cumulocity IoT. Accesses live OpenAPI specs for Core API and Digital Twin Manager (DTM).
Setup the local testing environment for Cumulocity applications to test UI and microservices.
Critical implementation patterns for building Cumulocity microservices that act as MCP (Model Context Protocol) servers. Use when developing MCP servers.
Guidelines for developing microfrontends, plugins, and blueprints in Cumulocity, including manifest configuration and package structure. Use when working with microfrontends.
Guidelines for what needs to be ensured if you want to create a new open source repository in the Cumulocity organization.
Standard practices for developing, testing, and building Python microservices in the Cumulocity ecosystem. Use when creating or refactoring Python microservices.
| name | c8y-ai-service |
| description | REST API documentation and guidance for interacting with the Cumulocity AI service. Use when sending messages to the AI agent. |
The Cumulocity AI Service provides a central manager for AI agents, LLM providers, and Model Context Protocol (MCP) servers.
Base URL: /service/ai
Use standard Cumulocity authentication (Basic Auth or OAuth2) in the Authorization header.
Text agents are used for conversational interactions. They can use tools and maintain history.
POST /agent/text/{name}
Parameters:
fullResponse (query, boolean): If true, returns detailed steps, tool usage, and token counts. Default is false (returns only text).Accept (header): application/json (default), text/event-stream (for SSE streaming), or application/x-ndjson.Request Body (TextCompletionDto):
{
"prompt": "Simple text question", // XOR messages
"messages": [
{"role": "user", "content": "Hello!"}
],
"system": "Optional system prompt override",
"temperature": 0.7,
"maxOutputTokens": 1024
}
GET /agent/text: List all available text agents.POST /agent/text: Create a new text agent. Requires AgentTextDefinitionDto.GET /agent/text/{name}: Retrieve agent configuration.PUT /agent/text/{name}: Update agent configuration.DELETE /agent/text/{name}: Delete an agent.Object agents generate structured JSON data based on a defined schema.
POST /agent/object/{name}
Request Body (ObjectCompletionDto):
{
"prompt": "Extract name and age from: John is 30.",
"schema": {
"type": "object",
"properties": {
"name": { "type": "string" },
"age": { "type": "number" }
},
"required": ["name", "age"]
}
}
GET /agent/object: List all available object agents.POST /agent/object: Create a new object agent.MCP servers provide tools that agents can use to interact with Cumulocity or external systems.
GET /mcp/tools
Returns a list of all tools available from all configured MCP servers.
GET /mcp/servers: List all configured MCP servers.POST /mcp/servers: Add a new MCP server (SSE or stdio).GET /mcp/servers/{name}/tools: List tools for a specific server.A global provider (e.g., OpenAI, Anthropic, Google) must be configured for agents to work.
GET /provider: Returns the current global provider (masked API key).POST /provider: Configure the global provider.AgentTextDefinitionDto{
"name": "agent-name",
"type": "text",
"agent": {
"system": "You are a helpful assistant.",
"temperature": 0.7
},
"availability": "PRIVATE" // or "SHARED"
}