Skip to main content
Glama
minghsuy

ctscout

by minghsuy

A vendor's customer counts, and — with a key — the customer enumeration

ctscout_vendor_customers
Read-onlyIdempotent

Get a vendor's customer summary or detailed per-customer list from CT logs. Choose the free summary for counts, or enumerate confirmed and candidate customers with an API key.

Instructions

Read the ctscout research product's vendor objects: the free summary for a vendor slug (GET /vendors/{slug}), or the per-customer enumeration (GET /vendors/{slug}/customers, which needs an API key). Debits no quota either way — both are precomputed objects published by the ctscout-research refresh.

Args:

  • slug (string, required): the vendor slug, one lowercase segment, e.g. 'cloudflare'. The values in a LEI record's vendors_confirmed are exactly these slugs.

  • enumerate (boolean, optional, default false): false = the free summary; true = the per-customer enumeration, which requires an active ctscout.dev API key (any tier) in CTSCOUT_API_KEY. Free returns up to 100 rows from each list; Pro returns the full published lists (which may still be export-capped or shortened for MCP). A missing, invalid or revoked key gets HTTP 401 and this tool explains that the summary is still available with enumerate: false.

  • response_format ('markdown' | 'json', default 'markdown'): output format.

Candidates and confirmed are two different claims and are NEVER summed:

  • Candidate = an apex domain this vendor certified a hostname for. Fan-out alone is not a vendor relationship: an organization certifying hundreds of its own product sites looks identical.

  • Confirmed = the DNS-confirmed subset of the candidates. The definition: a vendor is confirmed when a hostname it certified resolves onto a domain it certifies and the customer's own www does not, or another organization certifies the apex.

  • Confirmed is a SUBSET of candidates, so adding the two double-counts. The markdown keeps them in separate tables and the JSON in separate fields; report them apart.

Returns (on success, structuredContent follows the declared outputSchema; a failed call — 400, 401, 404, 503, timeout — is isError with no structuredContent):

  • enumerate: false (the summary): { "slug": string, "vendor_name": string, "vendor_apex": string | null, // null when the vendor's brand token matches no label it certifies "customers": { "candidates": number, "confirmed": number }, "countries_top": [ { "country": string, "confirmed": number } ], // CONFIRMED customers only "co_use": [ { "slug": string, "confirmed": number } ], // see below "sample_customers": [string] // hash-chosen sample of the CONFIRMED customers, whatever size the export published }

  • enumerate: true (the enumeration): { "slug": string, "confirmed": [ { "apex": string, "attributed_to": string | null, "lei": string | null } ], "candidates": [ same row shape ], "counts": { "candidates": number, "confirmed": number }, // what the research build holds "capped": boolean, // true = the build itself kept a subset of the candidates "truncated": boolean, // true = the API shortened a list for the key tier "free_slice": { "rows": number, "full_list": "pro" } | null, // Free limit per list; null for Pro "truncation_note": string // written by THIS server, only when it dropped rows to fit the character limit; counts and capped still describe the API's answer }

  • Both also carry "as_of" / "product_version", "snapshot_dates", and this server's "snapshot" / "snapshot_source" ("product" | "unavailable"; null snapshot means unknown freshness, never "current").

Reading the fields honestly:

  • co_use counts THIS vendor's confirmed customers that the other vendor also certifies — a candidate there, not a mutual confirmation.

  • countries_top counts confirmed customers that resolved to an LEI; candidates and LEI-less customers are not in it.

  • attributed_to is GLEIF's legal name when the apex resolves to one LEI, the single non-vendor certificate organization otherwise, and null when neither holds. It is an attribution, not an ownership claim.

Examples:

  • Use when: "How many customers does Cloudflare have in the index?" -> { slug: "cloudflare" } (report candidates and confirmed separately)

  • Use when: "List Cloudflare's confirmed customers" -> { slug: "cloudflare", enumerate: true }

  • Don't use when: you have a company and want its vendors — read vendors_confirmed from ctscout_lookup_lei instead.

Coverage & freshness:

  • A 404 means the slug is not in the published version, not that the vendor does not exist. The export is republished by the ctscout-research refresh, so these answers move on that cadence rather than the daily /scan warehouse sync.

  • Before the first publish the routes answer HTTP 503 and this tool returns a plain "not published yet" error. That is expected, not a fault in the query.

Corrections:

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
slugYesThe vendor's slug, e.g. 'cloudflare'. The slugs in a LEI record's vendors_confirmed are exactly these values.
enumerateNoOptional, default false. false returns the free vendor summary (counts, top countries, co-use, a customer sample). true returns the per-customer enumeration from GET /vendors/{slug}/customers, which requires an active ctscout.dev API key (any tier).
response_formatNoOutput format: 'markdown' for the counts (or the two customer tables), 'json' for the raw API response.markdown

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault
slugNoThe vendor slug the answer is filed under.
as_ofNoThe product version this answer was read from.
cappedNoEnumeration view: true when the research build kept a subset of the candidates. Says nothing about this server's own truncation — see truncation_note.
co_useNoSummary view: other vendors certifying this vendor's confirmed customers. The count is this vendor's confirmed customers that the other vendor also certifies — a candidate there, not a mutual confirmation.
countsNoEnumeration view: the completeness metadata — the rows the research build holds. counts.candidates can exceed candidates.length; see `capped`.
snapshotYesThe research product version (YYYY-MM-DD) this answer was read from — the `as_of` of the export the ctscout-research refresh published. null when the API response carried none.
confirmedNoEnumeration view: the DNS-confirmed customer rows.
customersNoSummary view: the candidate/confirmed split. Two different claims about the same vendor — confirmed is the DNS-confirmed subset of candidates, so adding them double-counts.
truncatedNoEnumeration view: the API shortened either list for this key's tier.
candidatesNoEnumeration view: the candidate customer rows. counts.confirmed is a subset of counts.candidates, but when `capped` is true the LISTED candidates are a hash-chosen subset that may omit rows the confirmed list carries.
free_sliceNoEnumeration view: free-tier rows allowed per list and the tier for the published list; null for Pro. Separate from export and MCP size caps.
vendor_apexNoSummary view: the vendor's own apex, null when its brand token matches none.
vendor_nameNoSummary view: the vendor's certificate name.
countries_topNoSummary view: top countries by CONFIRMED customers (candidates are not counted).
snapshot_datesNoPer-source provenance from the product manifest: elf, gleif, isin, psl, wikidata. Values are the dated snapshot each join read (psl is a bundle identifier, not a date).
product_versionNoSame value as as_of; the manifest's version.
snapshot_sourceYes'product' = the API response carried the export's as_of; 'unavailable' = it did not, snapshot is null and must be treated as unknown, never as current.
truncation_noteNoWritten by this MCP server, never by the API: present only when rows were dropped from the lists above to stay under the response character limit. `counts` and `capped` still describe the API's answer, not this list.
sample_customersNoSummary view: a hash-chosen sample of the CONFIRMED customers — whatever size the research export published, not a ranking and not a complete list.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed2 schema fields changedv0.6.1
    • addedOutput schema / properties / free_slice
      Added value: +{
      +  "anyOf": [
      +    {
      +      "additionalProperties": {},
      +      "properties": {
      +        "full_list": {
      +          "type": "string"
      +        },
      +        "rows": {
      +          "type": "number"
      +        }
      +      },
      +      "required": [
      +        "rows",
      +        "full_list"
      +      ],
      +      "type": "object"
      +    },
      +    {
      +      "type": "null"
      +    }
      +  ],
      +  "description": "Enumeration view: free-tier rows allowed per list and the tier for the published list; null for Pro. Separate from export and MCP size caps."
      +}
    • addedOutput schema / properties / truncated
      Added value: +{
      +  "description": "Enumeration view: the API shortened either list for this key's tier.",
      +  "type": "boolean"
      +}
  2. Addedv0.6.0

TDQS

A4.9/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

While annotations declare readOnlyHint, openWorldHint, idempotentHint, and destructiveHint, the description adds substantial behavioral context: it debits no quota, explains that a missing/invalid API key yields HTTP 401 and that the summary still works with enumerate: false, clarifies that 404 means the slug isn't in the published version, and that 503 means 'not published yet'. It also distinguishes candidates from confirmed and warns against summing them, and explains the 'capped' and 'truncated' flags. These go far beyond the annotations and are consistent with them.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is long but well-structured with a clear opening summary, sections for arguments, returns, examples, coverage, and corrections. It is front-loaded with the core purpose. However, it could be tightened; some redundancies exist (e.g., the return structure is described in both prose and the output schema). The length is justified by the tool's complexity, so it earns a 4 rather than lower.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The description covers every aspect an agent needs: the two modes, output structure (including nested fields and their meanings), error handling (400, 401, 404, 503, timeout), freshness semantics (as_of, product_version, snapshot), attribution rules for attributed_to, and the co_use semantics. It even includes examples and a corrections channel. Nothing critical is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Even though the schema provides descriptions for all three parameters, the description enriches each: it ties slug to the exact values in a LEI record's vendors_confirmed, details the free vs Pro limits and key requirements for enumerate, and explains the difference between markdown and json output for response_format. It also adds edge-case behavior like the 401 fallback, which the schema omits.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a precise statement: 'Read the ctscout research product's vendor objects: the free summary for a vendor slug (GET /vendors/{slug}), or the per-customer enumeration (GET /vendors/{slug}/customers, which needs an API key).' It names both modes and explicitly contrasts with a sibling in the 'Don't use when' note, directing to ctscout_lookup_lei for vendor lookups. This makes the tool's purpose unambiguous and distinct.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives explicit use cases with concrete examples: 'How many customers does Cloudflare have in the index?' and 'List Cloudflare's confirmed customers', plus a clear 'Don't use when' instruction that points to an alternative tool. It also explains when to set enumerate: true versus false, leaving no ambiguity about selection.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.