DOCUMENTATION / OVERVIEW

Agentis Documentation

Observe, evaluate, govern, and enforce policy on your autonomous AI agents in production with zero code changes.

Quickstart

Drop Agentis into any Python or TypeScript agent application:

quickstart.py

# 1. Install Agentis
pip install agentis

# 2. Add to your agent workflow
from agentis import init, observe

init(project="customer-support", redact_pii=True)

@observe(name="agent_reasoning", as_type="agent")
def handle_inquiry(prompt: str):
    # LLM calls, tool usage & costs are traced automatically
    return "Refund processed."

Explore Documentation

SDK Quickstart

Learn how to initialize Agentis, configure exporters, and observe functions with decorators.

Core Concepts

Understand agent lifecycle identity, distributed trace trees, spans, and OpenTelemetry semantics.

Framework Integrations

Guides for OpenAI Agents SDK, LangChain, CrewAI, AutoGen, and Google Gemini.

Compliance & Governance

EU AI Act Articles 12, 14, 50 compliance evidence exporters and HIPAA healthcare safeguards.