Tool description
Names, purposes, context and preconditions help callers choose correctly. A tool should have a specific job.

Explore enterprise products and AI capabilities built for industry.
Let’s talk about your businessProducts, integrations and partnershipsBring AI into the enterprise.
Automated data-warehouse builder
Automated high-quality datasets
AI appliance for business decisions
Opens in a new tabDedicated offline AI bidding appliance
Connect AI to products and operations.
Explore the interfaces and integration paths for our AI capabilities.
Models, foundational frameworks and the engineering behind AI systems.
Explore collaboration across products, technology and delivery.
Frameworks, tools and examples for developers to understand and reuse.
Updates, technical thinking and industry observations from IDENIFE.
Our focus on education, talent and sharing knowledge.
A clear tool description. Inputs that can be checked. Outputs another system can use. Connectors give systems a shared language for working together.
DISCOVEROne tool. One contract.
datasetread-onlyschema + metadataMCP Catalog Kit exposes the schemas of two synthetic datasets through one read-only tool. Change the dataset to inspect its arguments and result. The archive includes the same local example.
{
"name": "catalog.describe",
"arguments": {
"dataset": "orders"
}
}Only listed datasets are accepted. Extra arguments are rejected.
| Field | Type |
|---|---|
order_id | string |
region | string |
amount | number |
This preview uses local sample mappings, not your database. Tool hints do not replace authorization checks in a real system.
Moving from a local example to a business connector means specifying the responsibility of each call, not merely changing an address.
Names, purposes, context and preconditions help callers choose correctly. A tool should have a specific job.
Type checks are not authorization. Define access by identity, data scope and action; keep credentials in a controlled runtime.
Return understandable structures and distinguish invalid arguments, execution failures and unmet business conditions.
Unpack the archive and run the commands inside it. The example initializes, discovers a tool and reads a catalogue. Tests cover initialization, invalid arguments and unknown datasets.
node server.mjs < requests.jsonl
node --test server.test.mjsScope: stdio and one read-only tool. Production integration needs authentication, monitoring, version compatibility and business access control.
Read the full README ↗The reference is pinned to MCP 2025-11-25. These links point to that version of the official specification.