generated: '2026-07-19'
method: searched
source: https://docs.hydrolix.io/ + https://github.com/hydrolix/hdxcli + https://github.com/hydrolix/mcp-hydrolix
authentication:
  style: bearer
  detail: >-
    Credential login (username/password) exchanged for a bearer token, or a service-account
    token for automation. See authentication/hydrolix-authentication.yml.
  ref: authentication/hydrolix-authentication.yml
apis:
  config:
    role: Configure and manage the cluster (orgs, projects, tables, transforms, functions, jobs, service accounts).
    reference: https://docs.hydrolix.io/latest/openapi/config-api/
  stream:
    role: Real-time data ingestion (HTTP streaming ingest endpoint).
    reference: https://docs.hydrolix.io/latest/openapi/stream/
  query:
    role: Execute SQL queries and retrieve results (ClickHouse SQL dialect).
    reference: https://docs.hydrolix.io/latest/openapi/query/
query_language:
  dialect: ClickHouse SQL
  detail: >-
    Queries are ClickHouse-compatible SQL; the MCP run_select_query tool and the Trino/Spark/
    Splunk/Grafana connectors all read via the ClickHouse-compatible interface.
ingestion:
  modes: [HTTP streaming API, Amazon Kinesis, Data Firehose, S3 autoingest, OpenTelemetry (OTLP)]
  schema: Transforms define the write-time ingest schema per table.
idempotency:
  supported: unknown
  notes: No documented idempotency-key contract was found; not asserting one.
pagination:
  style: unknown
  notes: No public OpenAPI spec captured to confirm pagination parameters.
error_envelope:
  format: unknown
  notes: Not confirmed; no public spec captured.
rate_limiting:
  signaling: unknown
versioning:
  ref: lifecycle/hydrolix-lifecycle.yml