Skip to main content
Glama
fiware-orion-mcp

fiware-orion-mcp

by k2iser

fiware-orion-mcp

MCP (Model Context Protocol) server for FIWARE Orion Context Broker. Enables LLMs like Claude to manage NGSI-v2 context entities — IoT devices, cameras, sensors, vehicles, and any real-world object modeled as a digital twin.

No official FIWARE MCP exists yet. This is the first one.

What is FIWARE Orion?

FIWARE is an open standard (ETSI NGSI-LD) for Smart City and IoT data management. The Orion Context Broker is its core component — it stores the current state of all entities and notifies subscribers when things change.

Related MCP server: aas-mcp

Tools available

Tool

Description

orion_get_version

Verify Orion connectivity and version

orion_list_types

List all entity types and their attributes

orion_query_entities

Query entities by type, pattern, filter, geo

orion_get_entity

Get full details of a specific entity

orion_create_entity

Create a new NGSI entity

orion_update_attributes

Update entity attributes

orion_delete_entity

Delete an entity

orion_create_subscription

Subscribe to attribute change notifications

orion_list_subscriptions

List active subscriptions

quantumleap_get_history

Get time-series history for an entity

quantumleap_get_type_history

Get aggregated history across entity type

Quick start

1. Start FIWARE stack (Docker)

cd docker
docker compose up -d
# Verify:
curl http://localhost:1026/version
curl http://localhost:8668/version

2. Install MCP server

pip install fiware-orion-mcp
# or from source:
pip install -e .

3. Configure Claude Code

Add to your ~/.claude/mcp.json or project MCP config:

{
  "mcpServers": {
    "fiware-orion": {
      "command": "python",
      "args": ["-m", "fiware_mcp.server"],
      "env": {
        "ORION_URL": "http://your-orion-host:1026",
        "QUANTUMLEAP_URL": "http://your-ql-host:8668"
      }
    }
  }
}

4. Environment variables

Variable

Default

Description

ORION_URL

http://localhost:1026

Orion Context Broker URL

QUANTUMLEAP_URL

http://localhost:8668

QuantumLeap URL

FIWARE_SERVICE

(empty)

Multi-tenancy service name

FIWARE_SERVICE_PATH

/

Multi-tenancy service path

Example session with Claude

> List all entity types in the Context Broker
> Show me all Camera entities that are offline
> Get the full details of Camera:TIMONE-01
> Update Camera:TIMONE-01 status to "maintenance"
> Show me the uptime history of Camera:TIMONE-01 for the last 7 days

Roadmap

  • NGSI-LD support (Orion-LD)

  • Batch operations

  • Geo-query helpers (find entities near a point)

  • Export entities as GeoJSON

License

MIT — contributions welcome.

Available Tools

11 tools
orion_create_entityA

Create a new entity in the Context Broker.

Args: entity_id: Unique entity ID (e.g. 'Camera:CAM-001'). Convention: 'Type:Name'. entity_type: Entity type (e.g. 'Camera', 'TrafficSensor', 'ParkingSpot'). attributes: Dict of attribute definitions. Each attribute: {"attr_name": {"type": "Text|Number|Boolean|geo:json|DateTime", "value": }} Example: {"status": {"type": "Text", "value": "online"}, "temperature": {"type": "Number", "value": 22.5}}

Inputs: entity_id (required), entity_type (required), attributes (required). Returns: result

orion_create_subscriptionA

Create a subscription to receive notifications when entities change.

Args: entity_type: Entity type to watch (e.g. 'Camera'). watched_attrs: List of attribute names that trigger notification. notification_url: HTTP endpoint to receive POST notifications. id_pattern: Optional regex to filter by entity ID. description: Human-readable description.

Inputs: entity_type (required), watched_attrs (required), notification_url (required), id_pattern, +1 more. Returns: result

orion_delete_entityA
DestructiveIdempotent

Delete an entity from the Context Broker.

Args: entity_id: Entity ID to delete.

Inputs: entity_id (required). Returns: result

orion_get_entityA
Read-only

Get full details of a specific entity by its ID.

Args: entity_id: Entity ID (e.g. 'Camera:TIMONE-01'). attrs: Optional comma-separated list of attributes to return.

Inputs: entity_id (required), attrs. Returns: result

orion_get_versionA
Read-only

Get Orion Context Broker version and uptime. Use to verify connectivity.

Inputs: no named parameters. Returns: result

orion_list_subscriptionsA
Read-only

List all active subscriptions in the Context Broker. Subscriptions trigger HTTP notifications when entity attributes change.

Inputs: no named parameters. Returns: result

orion_list_typesA
Read-only

List all entity types registered in the Context Broker. Returns type names and attribute counts. Use this first to understand the data model.

Inputs: no named parameters. Returns: result

orion_query_entitiesA
Read-only

Query entities from the Context Broker.

Args: entity_type: Filter by type (e.g. 'Camera', 'TrafficSensor'). id_pattern: Regex pattern on entity ID (e.g. 'Camera:TIMONE.*'). q: Simple Query Language filter (e.g. 'status==online', 'temperature>25'). attrs: Comma-separated list of attributes to return (e.g. 'status,location'). limit: Max results (default 20, max 1000). offset: Pagination offset.

Inputs: entity_type, id_pattern, q, attrs, +2 more. Returns: result

orion_update_attributesA
Idempotent

Update one or more attributes of an existing entity.

Args: entity_id: Entity ID to update. attributes: Dict of attributes to update. Example: {"status": {"type": "Text", "value": "offline"}, "snmp_uptime": {"type": "Number", "value": 3600}}

Inputs: entity_id (required), attributes (required). Returns: result

quantumleap_get_historyA
Read-only

Get historical time-series data for an entity from QuantumLeap.

Args: entity_id: Entity ID (e.g. 'Camera:TIMONE-01'). entity_type: Entity type for disambiguation. attr: Specific attribute to retrieve (e.g. 'temperature'). All attrs if omitted. from_date: ISO8601 start date (e.g. '2026-03-01T00:00:00'). to_date: ISO8601 end date (e.g. '2026-03-07T23:59:59'). limit: Max data points (default 100). aggr_method: Aggregation: 'count', 'sum', 'avg', 'min', 'max'. aggr_period: Aggregation period: 'year', 'month', 'day', 'hour', 'minute', 'second'.

Inputs: entity_id (required), entity_type, attr, from_date, +4 more. Returns: result

quantumleap_get_type_historyA
Read-only

Get aggregated historical data for ALL entities of a given type. Useful for fleet-level analytics (e.g. average temperature across all sensors).

Args: entity_type: Entity type (e.g. 'Camera'). attr: Attribute to aggregate (e.g. 'snmp_uptime'). from_date: ISO8601 start date. to_date: ISO8601 end date. limit: Max data points. aggr_method: 'count', 'sum', 'avg', 'min', 'max'. aggr_period: 'year', 'month', 'day', 'hour', 'minute', 'second'.

Inputs: entity_type (required), attr (required), from_date, to_date, +3 more. Returns: result

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.1/5.0

    Scored across 11 tools

    Disambiguation5/5

    All 11 tools have clearly distinct purposes, with no overlap. Each tool targets a specific operation (create, read, update, delete, list, query) on distinct resources (entities, subscriptions, types, history).

    Naming Consistency5/5

    All tool names follow a consistent snake_case verb_noun pattern with clear domain prefixes (orion_ and quantumleap_), ensuring easy readability and predictable naming.

    Tool Count5/5

    11 tools is well-scoped for a Context Broker with historical data integration, covering essential CRUD, query, subscription, and history operations without unnecessary bloat.

    Completeness3/5

    The tool set covers core entity CRUD and queries, but lacks update and delete operations for subscriptions, which is a notable gap that may cause agent failures in subscription lifecycle management.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      D
      maintenance
      Connect your AI workflows to the ThingsBoard IoT Platform through this MCP server. Enables LLMs to query device telemetry, manage IoT entities (devices, assets, customers), and analyze sensor data - all through natural language. Perfect for building AI-powered IoT monitoring, predictive maintenance,
      99
      Apache 2.0
    • A
      license
      B
      quality
      D
      maintenance
      MCP server enabling AI assistants to perform full CRUD operations on Asset Administration Shells, Submodels, and Submodel Elements via the Eclipse BaSyx REST API.
      25
      34 PyPI
      7
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      MCP server that provides condensed and normalized data from OpenCTI to LLMs, enabling lookup of observables, adversaries, and reports with enriched context.
      16
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      MCP server for interacting with QUADS infrastructure systems via API, enabling resource management and automation through LLM applications.
      MIT