Provides LLM-as-judge and code-based evaluators for scoring LLM outputs, with built-in templates for hallucination, relevance, and toxicity detection.

Safe sandboxed shell execution for LegionIO

Five-level extinction protocol with archival, audit trail, governance gates, and configurable settings

Budget enforcement, cost attribution, and FinOps reporting for LegionIO

Connects LegionIO to Azure AI Foundry (model catalog, deployments, connections)

Connects LegionIO to Google Gemini API

Connects LegionIO to GitHub

Connects LegionIO to GitLab via the REST API

Used to connect Legion to Google Calendar

AIRB compliance gates and governance policy enforcement for LegionIO

Used to connect Legion to Grafana

Used to read heartbeats and updates the db

Used to connect Legion to Home Assistant

Connections LegionIO to any HTTP source

Used to connect Legion to influxdb

Used to connect Legion to Infoblox WAPI for DNS and network management

Connects LegionIO to Jenkins CI/CD

Connects LegionIO to JFrog products (Artifactory, Xray, Pipelines)

Used to connect Legion to Jira

Connects LegionIO to Kerberos/SPNEGO authentication and LDAP group resolution