# Validibot

> Validibot is an open-source, self-hosted data-validation platform. Build
> reusable workflows from JSON Schema, XML Schema, CEL, Tabular, SHACL,
> EnergyPlus, FMU, and other validators, then run submissions from the web UI,
> REST API, CLI, or optional MCP-connected AI agent. Runs produce structured
> findings and durable evidence.

Validibot is particularly useful as an independent checkpoint after
probabilistic AI generation. An agent can submit a candidate JSON document,
CSV, RDF graph, simulation model, or other technical artifact; use precise
findings to repair it; and release it only after the operator's workflow
passes. This guards against hallucinated structure and inconsistent data where
correctness can be expressed as schemas, rules, graph constraints, or
executable models.

Validibot is designed to run on infrastructure its operator controls. The
Community edition is AGPL-3.0 software, while commercial self-hosted editions
add collaboration, support, licensing, signed credentials, and optional MCP
integration. This file maps the user documentation at
https://docs.validibot.com/. A concatenated copy of every page is available at
https://docs.validibot.com/llms-full.txt.

## Getting started

- [Welcome](https://docs.validibot.com/): What Validibot is and where to start
- [Installation](https://docs.validibot.com/installation/): Run Validibot locally or on your own infrastructure
- [Getting started](https://docs.validibot.com/getting-started/): Your first workflow and validation run
- [Self-hosted editions](https://docs.validibot.com/self-hosted-editions/): Community, Pro, and Enterprise self-hosting

## AI-agent validation

- [Validate AI-generated data](https://docs.validibot.com/ai-agent-validation/): Put deterministic validation and bounded repair after probabilistic generation
- [Agent discovery](https://validibot.com/agents.txt): Deployment paths, permissions, licensing, and safe agent behaviour
- [Root product map](https://validibot.com/llms.txt): Concise machine-readable product and capability overview

## Core tasks

- [Manage workflows](https://docs.validibot.com/workflow-management/): Create workflows, add steps, define signals and assertions
- [Import and export workflows](https://docs.validibot.com/importing-and-exporting-workflows/): Move workflow definitions between deployments
- [Run validations](https://docs.validibot.com/running-validations/): Launch runs from the UI, API, or CLI
- [Review results](https://docs.validibot.com/reviewing-results/): Read findings, severities, evidence, and run reports
- [Collaborate and share](https://docs.validibot.com/collaboration/): Organizations, roles, and workflow sharing
- [Set up two-factor auth](https://docs.validibot.com/setting-up-mfa/): TOTP and recovery codes

## Validators

- [Validator overview](https://docs.validibot.com/validators-overview/): Compare built-in and advanced validation approaches
- [Basic and CEL](https://docs.validibot.com/basic-validator/): Deterministic project and cross-field rules
- [JSON Schema](https://docs.validibot.com/json-schema-validator/): Validate JSON against an explicit structural contract
- [XML Schema](https://docs.validibot.com/xml-schema-validator/): Validate XML against XSD
- [Tabular](https://docs.validibot.com/tabular-validator/): Validate CSV schemas and per-row CEL rules
- [SHACL](https://docs.validibot.com/shacl-validator/): Validate RDF and linked-data graphs
- [EnergyPlus](https://docs.validibot.com/energyplus-validator/): Run and check building-energy simulations
- [FMU](https://docs.validibot.com/fmu-validator/): Execute and validate Functional Mock-up Units

## API and agents

- [API overview](https://docs.validibot.com/api-overview/): Endpoints, deployment-local OpenAPI schema, pagination, and conventions
- [Data shapes and types](https://docs.validibot.com/api/data-shapes/): How payloads are represented
- [Authentication](https://docs.validibot.com/api/authentication/): API tokens and request authentication
- [Sending data to the API](https://docs.validibot.com/api/sending-data/): Raw-body, JSON-envelope, and multipart request modes
- [AI agents (MCP)](https://docs.validibot.com/api/mcp-integration/): Connect an MCP-capable agent to a self-hosted Pro deployment

## Reference

- [How data flows through a workflow](https://docs.validibot.com/concepts/data-namespaces/): Signals, step I/O, namespaces, lifecycle, and artifact boundaries
- [CEL expressions](https://docs.validibot.com/concepts/cel-expressions/): Assertion language and data namespaces
- [Glossary](https://docs.validibot.com/glossary/): Validibot terminology
- [Troubleshooting](https://docs.validibot.com/troubleshooting/): Common problems and fixes
- [FAQ](https://docs.validibot.com/faq/): Editions, licensing, and product questions

## Optional

- [Source code](https://github.com/mcquilleninteractive/validibot): Community repository and issue tracker
- [Operator documentation](https://github.com/mcquilleninteractive/validibot/tree/main/docs/operations/self-hosting): Production self-hosting guides