Call the same guardrails your compliance team manages from the DocsOrb SDK, the REST API, or the DocsOrb MCP server. Every check lands in AI Analytics.
@docsorb/sdk: docsorb.guardrails.evaluate()
REST API under /api/v1 with API keys
MCP server with OAuth for AI assistants

Rules get hardcoded
Each team writes its own filters, so the compliance policy and the running code drift apart.
Assistants run unchecked
Coding assistants and agents call tools and read data with no company rules in the loop.
No shared evidence
Checks in app code leave no trail compliance can review.

SDK and REST
Evaluate user input, tool calls, tool responses, and model output. Sanitize or anonymize personal data before it reaches the model.
MCP for AI assistants
Connect Cursor, VS Code, Claude Code, Claude Desktop, Codex, Gemini CLI, Windsurf, or ChatGPT with OAuth in a few clicks.
Keys and connected apps
Create API keys and see which MCP apps are connected from the Developers page.

Product teams shipping chat, agents, or AI features
A practical use case where this workflow gives teams cleaner ownership and proof.
Engineering orgs rolling out Cursor, Claude Code, or Codex
A practical use case where this workflow gives teams cleaner ownership and proof.
Platform teams that need one audit trail for AI checks
A practical use case where this workflow gives teams cleaner ownership and proof.

01
Start with your current policies, templates, owners, or approval process.
02
Define who needs to review, learn, acknowledge, or act on each update.
03
Keep versions, training, acknowledgments, and evidence connected in one place.
All databases and servers stay in Germany only. AI processing uses an EU-frontier LLM.
Support clear records for policies, training, and acknowledgments.
Built in Europe for teams that care about control and trust.
Common questions about this workflow.
Yes. DocsOrb is designed to help teams organize existing documents, improve ownership, and connect rollout activity around them.
No. It is useful for founders, HR, legal, operations, and security teams that need structured policy workflows without heavyweight tooling.
Policy content can feed into employee learning, comprehension checks, acknowledgment tracking, and evidence records.
Policy Management
Boards, versions, and publish.
Employee Portal
Assigned boards and Allowed AI.
Training
Flashcards, quizzes, certificates.
Acknowledgments
Version-locked campaign sign-off.
AI Inventory
Radar, use cases, risk, and owners.
AI Guardrails
Browser, MCP, SDK, and API.
Regulations
EU AI Act and ISO 42001 controls.
Ihre Privatsphäre ist uns wichtig
Wir verwenden Cookies, um DocsOrb sicher zu halten, die Nutzung der Website zu verstehen und Ihr Erlebnis zu verbessern. Sie können alle akzeptieren, optionale Cookies ablehnen oder Ihre Einstellungen verwalten. Datenschutzerklärung lesen