VectorFlow API

VectorFlow serves a REST API under /vectorflow/api/v1 on your install’s address, covering pipelines, the fleet, nodes and secrets. Its OpenAPI document is at /vectorflow/api/v1/openapi.json, and /vectorflow/api/v1/docs renders it as interactive docs.

The REST API takes a service-account key, which belongs to one environment. Create one under Settings → Service accounts and send it as Authorization: Bearer vf_live_….

/vectorflow/api/health/ready answers without a key, for load balancers and uptime checks.

Python and TypeScript clients

The veltro-sdk (Python) and @veltrosecurity/sdk (TypeScript) packages include a VectorFlow client. Point it at your address plus /vectorflow:

from veltro_sdk import BearerToken, VectorFlowClient

vectorflow = VectorFlowClient(
    "https://vectorflow.example.com/vectorflow",
    auth=BearerToken(service_account_key),
)
pipelines = vectorflow.pipelines.list()

import { BearerToken, VectorFlowClient } from "@veltrosecurity/sdk";

const vectorflow = new VectorFlowClient("https://vectorflow.example.com/vectorflow", {
  auth: new BearerToken(serviceAccountKey),
});
const pipelines = await vectorflow.pipelines.list();

Client SDKs covers installing them, timeouts and errors.

AI agents

VectorFlow’s tools are also available to AI agents over MCP at /mcp on the same address. See MCP.