MCP Prompts
Overview
MCP prompts are pre-configured request templates that guide your AI client to effectively use the platform's tools and resources.
Prerequisites
- API key with the MCP preset
- MCP client configured (see Configuration)
Reference
| Prompt | Description | Arguments |
|---|---|---|
explore-ontology | Discover and understand an ontology without knowing an identifier | question (optional) |
analyze-ontology | Analyze the structure of an ontology space | espaceId (required), focus (optional) |
query-ontology | Convert a natural language question into an ontology query | question (required), espaceId (required) |
create-workflow | Guide for designing a new workflow | description (required), trigger (optional) |
search-knowledge | Search the knowledge base with hybrid search | query (required), sourceId (optional), maxResults (optional) |
invoke-agent | Guide for invoking an AI agent with context | task (required), agentId (optional), context (optional) |
analyze-livedata | Analyze a connected data source | sourceId (required), focus (optional) |
explore-ontology
Discovers and explores an ontology without requiring a prior identifier.
Arguments
| Argument | Required | Description |
|---|---|---|
question | No | Question about the ontology (default: overview) |
Example
{
"jsonrpc": "2.0",
"method": "prompts/get",
"params": {
"name": "explore-ontology",
"arguments": {
"question": "Quelles entites sont modelisees ?"
}
},
"id": 1
}
Behavior: The prompt guides the AI client to use ontology_list_espaces then ontology_describe to explore the ontology automatically.
analyze-ontology
Analyzes the structure and relationships of an ontology space.
Arguments
| Argument | Required | Description |
|---|---|---|
espaceId | Yes | Identifier of the space to analyze |
focus | No | Specific aspect (types, relations, properties) |
query-ontology
Converts a natural language question into a structured query.
Arguments
| Argument | Required | Description |
|---|---|---|
question | Yes | Natural language question |
espaceId | Yes | Target ontology space |
create-workflow
Guides the design of a new workflow step by step.
Arguments
| Argument | Required | Description |
|---|---|---|
description | Yes | Description of what the workflow should do |
trigger | No | Trigger type (manual, schedule, webhook) |
search-knowledge
Searches the knowledge base with hybrid search.
Arguments
| Argument | Required | Description |
|---|---|---|
query | Yes | Question or search terms |
sourceId | No | Restrict to a specific source |
maxResults | No | Maximum number of results (default: 10) |
invoke-agent
Guides the invocation of an AI agent with the necessary context.
Arguments
| Argument | Required | Description |
|---|---|---|
task | Yes | Task description for the agent |
agentId | No | Identifier of a specific agent |
context | No | Additional context |
analyze-livedata
Analyzes the health, schema, and quality of a data source.
Arguments
| Argument | Required | Description |
|---|---|---|
sourceId | Yes | Identifier of the source to analyze |
focus | No | Specific aspect (schema, freshness, data quality) |
Limitations
- Prompts are guides -- the AI client may adapt the steps based on context.
- Some prompts require identifiers obtained through discovery tools.
Need Help?
Contact us: Support and Contact.