Getting started
The moving parts of DataNivra, what runs where, and the order in which to set things up.
Documentation
Developer and operator documentation
Reference material for engineers who install agents, call the API or operate the platform. For concepts, start with the learning center.
REST API overview
Authentication modes, error format, pagination and idempotency of the versioned v1 REST API.
Agent installation
Network, identity and secret-store requirements for running the customer-resident DataNivra agent.
Agent extension images for customer-installed drivers
How to add the IBM (Db2 and Informix), Teradata or SAP HANA driver, which DataNivra may not redistribute, to the agent by building a small extension image under the vendor's licence.
Agent protocol (datanivra.agent/v1)
The outbound lease protocol between agent and control plane, for security and network reviewers.
CLI
How the datanivra command line fits in, and why it can never become a raw-data tunnel.
Operator runbooks
Where operational runbooks live and the incident process for suspected raw-data egress.
Website lead data
How contact and demo requests from the public website are validated, stored, notified, retained and deleted.
SDKs and OpenAPI
The published OpenAPI 3.1 document, the typed Python and TypeScript SDKs, retries, pagination, idempotency and error handling — with examples.
Known limitations and residual risks
What DataNivra does not do yet, which risks remain after its controls, and the status of its legal documents, benchmarks and analytics.
User and administrator manuals
The seven DataNivra manuals — quick start, user, administrator, agent installation, security and architecture, enterprise deployment, and API/CLI/SDK — and how they stay true to the product.
Unstructured and AI test data (Preview)
Document sets, in-text masking with TEXT_SPAN_MASK, value-free document-set evidence, synthetic RAG evaluation sets and the default-deny model-provider policy — and their limitations.
Capacity planner
How the capacity and cost estimate works, which assumptions it uses, and why its numbers are estimates rather than measurements.
Test Data API overview
Approved test-data profiles that developers, CI/CD pipelines and AI agents invoke through REST, the CLI, YAML/JSON manifests and SDK helpers — with a five-minute synthetic quickstart.
Test-data profiles and manifests
Reference for TestDataProfile definitions and TestDataRequest invocations in YAML or JSON — every field, validation rules, pinned versions, exact counts, target pools and the automation envelope.
CI workload identity
Let a GitHub Actions or GitLab CI job use its own short-lived OIDC token instead of a person's session — trusted per repository, limited to approved profiles.
Decisions and automatic approval
How a test-data request is approved, sent to a person or denied — the decision rules, separation of duties, the reason codes, and the opt-in automatic approval envelope.
Test data in CI/CD pipelines
Request approved test data per pull or merge request, run your tests on a runner that reaches the approved target, and always release — GitHub Actions, GitLab CI and any compatible runner.
AI agents and MCP (Preview)
A hosted Model Context Protocol endpoint and a local adapter that let AI QA and coding agents discover, plan, request and release approved test data — bounded, metadata-only tools on the same authorized API.
Test data helpers in the Python and TypeScript SDKs
Manifest parsing and the plan, request, wait and always-release lifecycle in datanivra_sdk.testdata and @datanivra/sdk/testdata — not yet published; install from the repository or release assets.
Lease vs revoke and troubleshooting
What releasing a test-data request does (and never does), TTL and cleanup, request states, and how to fix denied, pending, exhausted and offline-agent situations.
API reference
Every v1 REST route, generated from the shared contracts.