API Access

The CRM exposes REST and GraphQL at the same base. REST is simpler for one-off CRUD; GraphQL is required for bulk operations and metadata.

Endpoints

GraphQL: https://crm.ancsports.net/graphql
REST base: https://crm.ancsports.net/rest
Metadata GraphQL: https://crm.ancsports.net/metadata

Auth

Workspace-scoped JWT. The key ANC scripts use lives in:

/root/rag2/services/integrations/twenty/crmAutomation.ts

All our bulk scripts read it from there. Request header:

Authorization: Bearer <JWT>
Content-Type: application/json

REST vs GraphQL — when to use each

Use REST	Use GraphQL
One record CRUD	Bulk create/update/delete
Per-record PATCH loops	Filter-based bulk update
Uploading files	Metadata queries/mutations
Favorites, reactions	Complex nested fetches

Common GraphQL snippets

Get opportunities by bidStatus

query {
  opportunities(filter: {bidStatus: {in: ["RFP_RECEIVED", "SCOPING"]}}) {
    edges { node { id name dealValue proposalDueDate } }
  }
}

Bulk update (filter-based)

mutation {
  updateOpportunities(
    data: {probability: 75},
    filter: {bidStatus: {eq: "SHORTLISTED"}}
  ) { affectedCount }
}

Bulk create

mutation {
  createOpportunities(data: [
    {name: "Deal A", companyId: "uuid", stage: "PROPOSAL"},
    {name: "Deal B", companyId: "uuid", stage: "PROPOSAL"}
  ]) { id }
}

Soft delete + restore

mutation { deleteCompanies(filter: {id: {in: ["uuid"]}}) { affectedCount } }
mutation { restoreCompany(id: "uuid") { id name } }

Company-merge cheat sheet

After identifying canonical + duplicate Company IDs:

# Reassign opportunities
mutation {
  updateOpportunities(
    data: {companyId: "CANONICAL"},
    filter: {companyId: {eq: "DUPLICATE"}}
  ) { affectedCount }
}

# Reassign people
mutation {
  updatePeople(
    data: {companyId: "CANONICAL"},
    filter: {companyId: {eq: "DUPLICATE"}}
  ) { affectedCount }
}

# Reassign venues
mutation {
  updateVenues(
    data: {companyId: "CANONICAL"},
    filter: {companyId: {eq: "DUPLICATE"}}
  ) { affectedCount }
}

# Reassign service tickets (note: companyIdId, not companyId)
mutation {
  updateServiceTickets(
    data: {companyIdId: "CANONICAL"},
    filter: {companyIdId: {eq: "DUPLICATE"}}
  ) { affectedCount }
}

# Soft-delete the duplicate
mutation {
  deleteCompanies(filter: {id: {in: ["DUPLICATE"]}}) { affectedCount }
}

DesignRequest / Estimate / RfpAnalysis don't have direct Company relations — they chain via Opportunity. Don't try to merge them through company.

REST: favorites

POST /rest/favorites
Content-Type: application/json

{
  "forWorkspaceMemberId": "uuid",
  "position": 0,
  "dashboardId": "uuid"  // or viewId, or opportunityId
}

Metadata queries

List all objects:

query { objects { id nameSingular namePlural isCustom } }

List all fields on Opportunity:

query {
  object(input: {id: "c779922d-cf25-4a5e-9382-23eb1c02199e"}) {
    fields { id name type options { value label } }
  }
}

List all agents (Boyka):

query { agents { id name prompt isCustom } }

Uploading files (multipart)

FILES fields accept multipart uploads via uploadFilesFieldFile:

curl https://crm.ancsports.net/metadata \
  -H "Authorization: Bearer $JWT" \
  -F 'operations={"query":"mutation($input: UploadFilesFieldFileInput!) { uploadFilesFieldFile(input: $input) { name fileId } }","variables":{"input":{"recordId":"uuid","fieldName":"generatedImage","files":[null]}}}' \
  -F 'map={"0":["variables.input.files.0"]}' \
  -F '0=@image.jpg'

This is how Designer AI uploads generated images.

Python client pattern

Our bulk scripts (at /root/rag2/scripts/) use this shape:

import httpx
JWT = open("/root/.twenty-jwt").read().strip()
client = httpx.Client(
    base_url="https://crm.ancsports.net",
    headers={"Authorization": f"Bearer {JWT}"},
    timeout=30.0,
)
r = client.post("/graphql", json={"query": "...", "variables": {...}})
r.raise_for_status()

API Access

On this page