> ## Documentation Index > Fetch the complete documentation index at: https://docs.teamfollowup.ai/llms.txt > Use this file to discover all available pages before exploring further. > ## Agent Instructions > Use the public API base URL: https://api.teamfollowup.ai/api. > Authenticate public API requests with Authorization: Bearer YOUR_API_KEY. > Use product-owned terms: agents, campaigns, projects (GoHighLevel sub-accounts), contacts, calls, outcomes, skills, cadences, phone numbers. > Scope requests by locationId rather than projectName. > Read the guide linked from each API reference group before recommending an endpoint. # Get split test > Read the Agent Split Test on the live campaign bound to this Primary Variant. Answers 404 when the agent has no live campaign, or when that campaign has never been converted to a split test. Required API key scope: `agents:read`. ## OpenAPI ````yaml /openapi.yaml get /api/agent-builder/agents/{id}/split-test openapi: 3.1.0 info: title: TFUAI Server API version: 1.0.0 description: >- API reference for the TFUAI SaaS server. All published endpoints are mounted under `/api`. Responses use a standard envelope: `{ "success": true, ... }` on success and `{ "success": false, "error": "..." }` on failure. Product API modules are being launched module by module. This spec currently exposes only contracted public modules. Authenticate public API requests with an API key in the `Authorization` header: `Authorization: Bearer YOUR_API_KEY`. servers: - url: https://api.teamfollowup.ai description: Production API origin. Public API paths are under /api. security: [] tags: - name: Agents description: >- Build the conversation: identity and settings, script, lead context and split testing. - name: Analytics description: >- Understand call volume, pickup, booking and transfer metrics within the selected reporting scope. - name: Billing description: >- Balances, spend, purchases and sub-account rebilling. Reference and guides only; the builder chatbot has no billing playbook. - name: Calls description: >- Review completed calls, recordings, transcripts and campaign run history; distinguish observation from placing a new call. - name: Campaigns description: >- Configure what starts calls, who is eligible, calling windows, caller IDs, activation and the actions that follow each outcome. What runs after a call lands: for each outcome (booked, opted out, asked for a callback…), an ordered chain of actions — tag the contact, move a pipeline stage, send an SMS, book an appointment. Stored as a graph: `nodes` are the actions, `edges` say what follows what, and `dispositionEntries` maps each outcome to the node its chain starts at. **Pick the narrowest endpoint that does the job.** | To | Call | | | --- | --- | --- | | See what fires per outcome | `GET .../workflow?view=summary` | Reads as plain text, no graph walking | | Add steps to an outcome | `POST .../outcomes/{outcome}/nodes` | Server derives node ids, edge ids, handles, layout | | Change a message, tag or setting | `PATCH .../workflow` | Merges; carries no graph, so it cannot damage one | | Remove one step | `DELETE .../nodes/{nodeId}` | Also re-links the chain around it | | Author or replace the whole graph | `PUT .../workflow` | Replaces everything you send | `PUT` is the only one that can re-wire, position nodes, build an `if`, save an unconfigured draft, or copy a whole workflow — and the only one that can overwrite a change someone else made after you read it. Use it when you are genuinely authoring the graph, and send `expectedVersion` when you do. For everything else the narrower endpoints are both easier and safer. - name: Contacts description: >- Find contacts and their calls, manage do-not-call status and identify internal contacts excluded from reporting. - name: In-call Capabilities description: >- Configure actions and instructions the agent uses during a live call: booking, transfers, lookups and custom skills. - name: Master agent description: >- Run a shared agent and campaign across linked projects; manage inheritance, local bindings, synchronization and rollout readiness. - name: Native Calendar description: >- The diary of a sub-account with no CRM: its bookable calendars, the free slots they offer, and the appointments on them. Availability is computed per request against live bookings and is never cached, because a cached slot is one somebody else has already taken. - name: Native Contacts description: >- The people a sub-account with no CRM calls, and the fields it keeps on them. This is the system of record for those contacts: there is nowhere else they exist. A sub-account with a CRM keeps its contacts there and every operation here answers 400 for it. - name: Phone Numbers description: >- Find and manage caller IDs, assign numbers to campaigns and understand connected Twilio numbers. - name: Projects description: >- Understand project ownership, inspect configuration and validate the local resources used by a campaign. - name: TFU Live description: >- Build one TFU Live agent with its Brain, Voice and equipped actions; connect it through shared Campaigns. - name: Voices description: >- Browse the voice catalogue and select a voice; configure speech and listening behavior in Agents. paths: /api/agent-builder/agents/{id}/split-test: get: tags: - Agents summary: Get split test description: >- Read the Agent Split Test on the live campaign bound to this Primary Variant. Answers 404 when the agent has no live campaign, or when that campaign has never been converted to a split test. Required API key scope: `agents:read`. operationId: get-agent-split-test parameters: - name: id in: path required: true schema: type: string minLength: 1 description: >- Agent id of the Primary Variant. Every split operation is addressed through the primary, not through a secondary variant. example: agent_8a488fcfa8fdbf8ce8ad5ccc45 responses: '200': description: The current split. content: application/json: schema: $ref: '#/components/schemas/SplitTestResponse' examples: default: value: success: true splitTest: locationId: loc_9f7a123 campaignId: campaign_speed_to_lead_123 primaryAgentId: agent_8a488fcfa8fdbf8ce8ad5ccc45 status: active revision: 4 sync: status: healthy failedAt: null variants: - agentId: agent_8a488fcfa8fdbf8ce8ad5ccc45 name: Speed to Lead sequence: 1 weight: 50 status: active addedAt: '2026-08-14T09:12:44.318Z' removedAt: null isPrimary: true - agentId: agent_7c4e1b28f0a94d6591cc2fd340 name: Speed to Lead II sequence: 2 weight: 50 status: active addedAt: '2026-08-20T16:04:02.771Z' removedAt: null isPrimary: false '400': description: Validation failed or the request is malformed. content: application/json: schema: $ref: '#/components/schemas/SplitTestErrorResponse' examples: default: value: success: false error: VALIDATION message: Weights must include every active variant exactly once. '401': description: Missing or invalid authentication. content: application/json: schema: $ref: '#/components/schemas/SplitTestErrorResponse' examples: default: value: success: false error: Unauthorized message: Authentication required. '403': description: >- Authenticated but missing the required scope or access to this agent. content: application/json: schema: $ref: '#/components/schemas/SplitTestErrorResponse' examples: default: value: success: false error: AccessDenied message: You do not have access to this agent. '404': description: >- No agent, no live campaign bound to it, or no Agent Split Test on that campaign. content: application/json: schema: $ref: '#/components/schemas/SplitTestErrorResponse' examples: default: value: success: false error: NOT_FOUND message: This campaign does not have an Agent Split Test. '429': description: Rate limit exceeded. content: application/json: schema: $ref: '#/components/schemas/SplitTestErrorResponse' examples: default: value: success: false error: TooManyRequests message: Too many requests, please try again later. '500': description: Unexpected server error. content: application/json: schema: $ref: '#/components/schemas/SplitTestErrorResponse' examples: default: value: success: false error: InternalError message: Unexpected server error. security: - ApiKeyAuth: [] x-codeSamples: - lang: bash label: cURL source: |- curl --request GET \ --url https://api.teamfollowup.ai/api/agent-builder/agents/{id}/split-test \ --header 'Authorization: Bearer YOUR_API_KEY' components: schemas: SplitTestResponse: type: object additionalProperties: false required: - success - splitTest properties: success: type: boolean splitTest: $ref: '#/components/schemas/SplitTest' SplitTestErrorResponse: type: object additionalProperties: {} required: - success - error - message properties: success: type: boolean error: type: string description: >- Machine-readable error code. `VALIDATION`, `NOT_FOUND`, `CONFLICT` and `SYNC_FAILED` (502: the agent behind a variant could not be renamed, and nothing was committed) are the ones this module raises. example: CONFLICT message: type: string description: Human-readable explanation safe to show to an operator. example: Agent Split Test changed since it was loaded. Refresh and retry. SplitTest: type: object additionalProperties: false required: - locationId - campaignId - primaryAgentId - status - revision - variants properties: locationId: type: string description: Project the campaign belongs to. example: loc_9f7a123 campaignId: type: string description: Campaign whose traffic is being split. example: campaign_speed_to_lead_123 primaryAgentId: type: string description: >- Agent id of the Primary Variant, and the id every split operation is addressed through. example: agent_8a488fcfa8fdbf8ce8ad5ccc45 status: type: string enum: - active - ended description: >- A split ends automatically when removal leaves one variant, which then takes all the traffic. example: active revision: type: integer minimum: 1 description: >- Increments on every change. Send the value you read back on any mutation; a mismatch answers 409 rather than overwriting somebody else's edit. example: 4 sync: $ref: '#/components/schemas/SplitTestSync' variants: type: array description: >- Every variant, in stable sequence order: the live ones, the ones excluded from the cycle (`splitDraft: true`), and the removed ones. items: $ref: '#/components/schemas/SplitTestVariant' SplitTestSync: type: object additionalProperties: {} required: - status description: >- Whether the split is in step with the external voice platform. `error` means a shared-field update could not be rolled back and traffic has been forced to the Primary Variant. properties: status: type: string enum: - healthy - error example: healthy failedAt: type: - string - 'null' format: date-time description: When the sync failure was recorded. Null while healthy. example: null lastError: type: - object - 'null' additionalProperties: {} description: >- Diagnostic detail for the failure. Shape is not contractual and is present only after a failure. SplitTestVariant: type: object additionalProperties: false required: - agentId - name - sequence - weight - status - isPrimary description: >- One arm of the split. The Primary Variant is the agent the campaign is bound to; the others are clones created by this API. properties: agentId: type: string description: >- Agent id for this variant. Use it with the ordinary agent endpoints to read or edit the variant itself. example: agent_7c4e1b28f0a94d6591cc2fd340 name: type: string description: >- Display name. By default a secondary variant is named after the primary with a Roman numeral suffix and follows a primary rename. A name chosen through `PATCH /split-test/variants/{variantAgentId}` stays until it is reset; automatic synchronization updates only the names still on the convention. example: Speed to Lead II customName: type: boolean description: >- Present and true when an operator chose this name. The convention leaves it alone; reset it with `PATCH /split-test/variants/{variantAgentId}` and `name: null`. Never set on the Primary Variant, whose name is the agent's own. example: false sequence: type: integer minimum: 1 description: >- Stable position used to derive the canonical name. Never reused, so a removed variant does not free its numeral. example: 2 weight: type: integer minimum: 0 maximum: 100 description: >- Share of campaign traffic, as a percentage. Weights across the live variants total 100. A variant excluded from the cycle (`splitDraft: true`) is 0, and so is a removed one. example: 50 splitDraft: type: boolean description: >- Present and true when this variant has been excluded from the split cycle. It stays a full member (shared post-call fields and campaign-owned tools still reach it, it can still be renamed, and a test call can still be sent to it) but it takes no calls and holds a `weight` of 0. Leads it was already speaking to are not lost: while it is excluded, each of their calls goes to one live variant, the same one every call, and nothing about the lead is rewritten, so including the variant again hands those leads straight back. Those covering calls do not count in the covering variant's own numbers. Absent means the variant is live. Never set on the Primary Variant, which is the fallback for every route and always takes traffic. example: false status: type: string enum: - active - removed description: >- Membership, not traffic: an excluded variant is still `active` and carries `splitDraft: true`. Removed variants stay in the list so historical call attribution keeps resolving. example: active addedAt: type: - string - 'null' format: date-time example: '2026-08-20T16:04:02.771Z' removedAt: type: - string - 'null' format: date-time description: Set when the variant was removed, null while active. example: null isPrimary: type: boolean description: >- True for the campaign's bound agent. The Primary Variant cannot be removed. example: false securitySchemes: ApiKeyAuth: type: http scheme: bearer bearerFormat: API key description: 'Send your API key as `Authorization: Bearer YOUR_API_KEY`.' ````