get_vendor_information next channel

Status: Modified (upstream-proxy endpoint migration; snake_case params; vendor_company_id added) This tool is on the private next MCP channel for allow-listed orgs. When promoted, this file moves to mcp-tools/get-vendor-information.md.

Search for vendors in the HG Insights taxonomy and retrieve their vendor_ids and metadata. Free — no credits consumed.

Integration required: hginsights_v2__data_api Cache schema version: 3 (differs from base — stale cached payloads from the public channel will not be served)

Parameters

Parameter	Type	Required	Default	Description
vendor_name	string	Optional	—	Case-insensitive substring match on vendor_name (LIKE). Activates relevance ranking. Use to search vendors by company name.
description	string	Optional	—	Case-insensitive substring match on vendor_company_description. ANDed with vendor_name when both are provided. If results are empty when using both filters, retry with only vendor_name — the stored description may not contain your exact phrase.
vendor_id	integer	Optional	—	Exact vendor_id (UInt64) match — returns ≤1 row. Use when you already know the ID and want to resolve full details.
has_products_with_installs	boolean	Optional	—	true = only vendors with ≥1 product carrying an install signal. Omit to return all. Note: product ownership joins may occasionally surface unrelated vendors — verify vendor_name and vendor_url before using the returned vendor_id.
include_products	boolean	Optional	false	When true, each vendor row carries a products[] of {product_id, product_name} ordered by presence frequency, capped at products_limit.
products_limit	integer	Optional	10	Cap on the products[] list per vendor when include_products is true (1–100).
sort_by	"relevance" | "vendor_name" | "product_count"	Optional	"relevance"	Sort order. On next, unknown values return HTTP 422 (previously fell back to default silently).
limit	integer	Optional	10	Maximum number of vendor rows to return (1–100).
offset	integer	Optional	0	Pagination offset.

Response Format

{
  "vendors": [
    {
      "vendor_id": 316,
      "vendor_name": "Snowflake Inc.",
      "vendor_url": "snowflake.com",
      "vendor_parent_id": null,
      "vendor_company_id": "3AB6196C…",
      "vendor_company_description": "Cloud data platform…",
      "product_count": 12,
      "products": [
        { "product_id": 26434, "product_name": "Snowflake" }
      ]
    }
  ],
  "total": 1,
  "has_more": false,
  "credits_consumed": 0
}

New field: vendor_company_id

vendor_company_id (string | null) is a new field in the next-channel output schema. It holds the paired HG company id for this vendor, when known. This field was not present in the base tool's response.

Behavioural differences on next

Aspect	Base tool (public)	next channel
sort_by validation	Unknown value silently falls back to default	Unknown value → HTTP 422 from upstream
Upstream error code	Non-retryable errors → HTTP 503	Non-retryable errors → HTTP 502
has_more source	Returned directly by the upstream envelope	Computed client-side: (offset + data.length) < count

Example Usage

{
  "tool": "get_vendor_information",
  "parameters": {
    "vendor_name": "Snowflake",
    "include_products": true,
    "products_limit": 5
  }
}

Related Tools

• company_technographic — filter installs by vendor_ids from here.
• company_fai — filter FAI rows by vendor_ids.
• company_install_time_series — filter time series by vendor_ids.

Changelog

2026-07-31 — Breaking: params renamed to snake_case

Availability: live on the private next MCP channel for allow-listed orgs. The public /api/mcp contract is unchanged.

	Before (on next)	After
Name filter	vendorName	vendor_name
ID filter	vendorId	vendor_id
Install filter	hasProductsWithInstalls	has_products_with_installs
Product include	includeProducts	include_products
Products cap	productsLimit	products_limit
Sort	sortBy	sort_by

description, limit, and offset were already snake_case and are unchanged. Output fields were already snake_case — no migration needed for output parsing.

2026-07-30 — Internal: endpoint migration to upstream-proxy v2

Availability: live now on the private next MCP channel for allow-listed orgs. The public /api/mcp contract is unchanged.

Why this changed: get_vendor_information now calls the upstream-proxy endpoint POST /data-api/v2/products/vendors/search (returns {count, data}) instead of the deprecated ClickHouse-backed POST /v2/products/vendors/search (returned {results, pagination, credits_consumed}). The response shape visible to callers is identical — the same vendors / total / has_more / credits_consumed envelope — but two low-level behavioural details change (see table above).

New output field: vendor_company_id (string | null) — paired HG company id for this vendor, when known. This field is new in the next-channel; it was not present in the base tool's response schema.

Cache schema version: getCacheSchemaVersion() returns super.getCacheSchemaVersion() + 2 (+2 to account for two independent breaking changes: the endpoint switch and the addition of vendor_company_id).

What stayed the same: all parameters and their behavior (including the "at least one of vendor_name / description / vendor_id required" validation rule), response shape, credit cost (free), and required integration (hginsights_v2__data_api).