get_entity
Fetch an AnchorID (canonical entity) by UUID. Optionally include linked source records via the include parameter (links, source_records, or both).
LABSCO SUMMARY
A three-state answer — resolved, needs_review, not_found — so an agent has somewhere to stop.
The value is in the middle state. When several candidates match with similar confidence the answer is needs_review with the candidates attached, which is the signal to surface them to a person instead of writing to the wrong record. guard_write extends that to the moment before a write: it evaluates and returns allowed or blocked with reasons, and performs no write itself, so the responsibility for acting stays where it belongs. Every response carries a request_id for tracing a decision back afterwards.
What it is
An identity resolution service reached over MCP. Companies and people resolve to a canonical AnchorID, source records link to it, and a pre-write check says whether a proposed write is safe. It is a thin proxy — matching and scoring happen on the hosted API.
What you get
- A company resolved from domain, name, city and state, or external identifiers, returning resolved, needs_review or not_found with a confidence score, the AnchorID, the reasons it matched, and any ambiguous candidates
- A person resolved the same way from email, name, company domain or external identifiers such as Slack and Google user IDs
- Both in batches of up to 200 per call, each item carrying a client_ref so results correlate back to your rows
- An entity fetched by UUID, optionally with its links and source records, and its golden record exported as one merged JSON view
- Source records linked to an AnchorID and unlinked again — both idempotent, so repeating the call returns the existing state rather than duplicating
- A pre-write check that verifies the AnchorID exists, that confidence meets your threshold, that there are no unresolved conflicts, and that at least one canonical link is present — returning allowed or blocked with reasons, and writing nothing itself
- The same check in batches of up to 200
- A source record ingested and matched to an AnchorID automatically, against a system you have registered
Requirements
An Anchord account and an API key from app.anchord.ai — created under Settings > API Keys and shown once. It is a Bearer token and every request is scoped to its tenant. Two ways to connect: locally with ANCHORD_API_KEY set and npx -y @anchord/mcp-server over stdio, from the npm package @anchord/mcp-server, currently 1.1.1, needing Node.js 20 or newer; or the hosted endpoint https://mcp.anchord.ai/mcp over streamable HTTP with the key in an Authorization: Bearer header and nothing installed at all. ANCHORD_API_BASE_URL overrides the default https://api.anchord.ai. A Dockerfile is included. Anchord never writes to your CRM or database — the write decision stays with the caller.
Setup effort
One command plus a key — ANCHORD_API_KEY=<YOUR_ANCHORD_API_KEY> npx -y @anchord/mcp-server
USE CASES
01
Checking a lead is not already in the CRM under another spelling
Resolve the company from its domain and name; ambiguous matches come back as needs_review with the candidates.
02
Deduplicating an imported list before it lands
Resolve up to 200 rows in one call, each tagged with your own reference so results map back.
03
Stopping an agent writing to the wrong record
Run the pre-write check first — it returns allowed or blocked with reasons and writes nothing.
04
Seeing everything known about one entity
Fetch it with its links and source records, or export the merged golden record as a single object.
05
Getting a new system's records into the graph
Ingest the record against its registered system and let it be matched to an AnchorID.
06
Correcting a record attached to the wrong entity
Unlink it from the wrong AnchorID and link it to the right one — both calls are idempotent.
11 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
get_entity_export
Export the golden record for an AnchorID. Returns the merged/canonical view of all linked source records as a single JSON object.
guard_write
Evaluation-only pre-write safety check. Verifies the AnchorID exists, confidence meets threshold, no unresolved conflicts, and at least one canonical link is present. Returns allowed/blocked with reasons. This tool does NOT perform any write — the caller decides whether to proceed.
guard_write_batch
Batch pre-write safety check for multiple AnchorIDs (max 200). Each item needs a client_ref for correlation. Returns per-item allowed/blocked decisions with reasons. Evaluation-only — never writes.
ingest_record
Ingest a single source record into Anchord. The record is matched to an AnchorID automatically. Requires a registered source (system). Wraps POST /ingest/batch with a single-item array.
link_source_record
Create or reactivate a link between an AnchorID and a source record. Idempotent — calling twice returns the existing link.
resolve_company
Resolve a company to an AnchorID using domain, name, city/state, or external identifiers. Returns status (resolved | needs_review | not_found), confidence score, the canonical AnchorID, match reasons, and any ambiguous candidates.
resolve_company_batch
Resolve multiple companies to AnchorIDs in a single call (max 200). Each item needs a client_ref for correlation and at least one identifying field. Ambiguous matches return status needs_review with candidate AnchorIDs.
resolve_person
Resolve a person to an AnchorID using email, name, company domain, or external identifiers (Slack/Google user IDs). Returns status (resolved | needs_review | not_found), confidence score, the canonical AnchorID, match reasons, and any ambiguous candidates.
resolve_person_batch
Resolve multiple people to AnchorIDs in a single call (max 200). Each item needs a client_ref for correlation and at least one identifying field. Ambiguous matches return status needs_review with candidate AnchorIDs.
unlink_source_record
Soft-delete the link between an AnchorID and a source record. Idempotent — calling on an already-unlinked pair returns 200.
BEFORE YOU INSTALL
What to know first
Tools you get
11 tools when we connected. The hosted endpoint exposes the same 11 tools as a local install.
Requires
- npx on PATH
- ANCHORD_API_KEY in your environment
Where it runs
Of the 3 documented clients, 2 run the server on your machine and 1 connects to a hosted endpoint.
CONTEXT COST
Light Standing cost
Rated by the size of the 11 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
Server setup
The install command the project publishes
TERMINAL
ANCHORD_API_KEY=<YOUR_ANCHORD_API_KEY> npx -y @anchord/mcp-serverHosted endpoint
ENDPOINT URL
https://mcp.anchord.ai/mcpAdd this hosted endpoint to your MCP client using the credentials listed under Requirements.
MCP Clients
Cursor, Claude Desktop
JSON
{
"mcpServers": {
"anchord": {
"command": "npx",
"args": [
"-y",
"@anchord/mcp-server"
],
"env": {
"ANCHORD_API_KEY": "<YOUR_ANCHORD_API_KEY>"
}
}
}
}Any MCP client — HTTP
ANY MCP CLIENT — HTTP · JSON
{
"mcpServers": {
"anchord": {
"url": "https://mcp.anchord.ai/mcp",
"headers": {
"Authorization": "Bearer <YOUR_ANCHORD_API_KEY>"
}
}
}
}
