Skip to documentation
Maetra
Documentation
Introduction
Getting started
Quickstart
Authentication
Errors & rate limits
Agents
Govern API
Checkpoints
Policies
Decision tokens
Execution receipts
Secure API
Rules
Scanning content
Incidents
Task Guard API
Tasks & context
Alignment & effects
MCP server
Overview & connection
Tools reference
End-to-end example
Reference
API reference
API v1
Public documentation
Maetra
Docs
Search documentation
⌘
K
Suggested
AI agent governance API and MCP documentation
Build safer AI agents with Task Guard, approval checkpoints, one-use execution authorization, effect receipts, runtime content scanning, and 18 MCP tools.
Overview
Getting started
Authenticate with Maetra, make your first API calls, and understand errors, rate limits, and idempotency.
Getting started
Quickstart
Go from an API key to a live approval checkpoint, decision-intelligence-ready policy evaluation, and content scan.
Getting started
Authentication
Create, scope, verify, and revoke Maetra workspace API keys for REST and MCP integrations.
Getting started
Errors and rate limits
Handle Maetra API errors, rate-limit responses, polling etiquette, and idempotent retries safely.
Getting started
Agents
Use registered or free-text agent identities in Maetra and list the agents available to your workspace.
Agents
Govern API
Route consequential AI agent actions through policy and human approval, bind decisions to exact payloads, and verify execution effects.
Govern API
Checkpoints
Create checkpoints for exact policies or decision intelligence, follow their lifecycle, and handle terminal outcomes.
Govern API
Policies
List active exact and decision-intelligence policies, and scope checkpoints to policies or policy groups.
Govern API
Maetra.io
Choose language
◎
EN
DE
FR
ES
Start free
Docs
Copy page
Secure API
Scan prompts, tool calls, and outputs against your rules, and review incidents.
Rules
List, create, and update Maetra Secure rules for prompts, data patterns, tools, and policy conditions.
Scanning content
Scan agent inputs, tool calls, and outputs and act on safe, flagged, or blocked verdicts.
Incidents
List security incidents raised by content scans and correlate them with the rules and agents involved.
Previous
Execution receipts
Next
Rules
