Skip to main content
Hexr is the runtime platform that gives every AI agent a cryptographic identity, authenticated cloud access, and full observability — automatically, with no infrastructure expertise required. Add @hexr_agent to any Python function, run three commands, and your agent is deployed to Kubernetes with mTLS networking, SPIFFE identity, and OpenTelemetry tracing out of the box.

Quick Start

Deploy your first agent in under 5 minutes. Install the SDK, write your agent, and run three commands.

Hexr Cloud

Fully managed SaaS on GKE. Sign up, get an API key, and deploy agents without managing infrastructure.

SDK Reference

Complete reference for @hexr_agent, hexr_tool, hexr_llm, and all SDK modules.

CLI Reference

Reference for hexr build, hexr push, hexr deploy, and all CLI commands.

​
From decorator to production in three commands

1

Install the SDK

pip install "hexr-sdk[cli]" --extra-index-url https://pypi.hexr.cloud/simple/
2

Write your agent

from hexr import hexr_agent, hexr_tool, hexr_llm
import openai

@hexr_agent(name="research-analyst", tenant="acme-corp")
def analyze(topic: str):
    s3 = hexr_tool("aws_s3")            # authenticated — no keys in code
    client = hexr_llm(openai.OpenAI())   # auto-traced with cost attribution
    return client.chat.completions.create(
        model="gpt-4o",
        messages=[{"role": "user", "content": f"Analyze {topic}"}]
    ).choices[0].message.content
3

Build, push, and deploy

hexr build   # AST scan → Dockerfile + Kubernetes manifests
hexr push    # Container image → registry
hexr deploy  # Schedule on Kubernetes → SPIFFE identity → ready

​
What you get automatically

Every Hexr agent receives all of the following with zero configuration:

Cryptographic Identity

Every agent process gets a unique SPIFFE X.509 certificate. Mutual TLS on every connection. Certificates rotate automatically.

Zero-Secret Cloud Access

Call hexr_tool("aws_s3") and get an authenticated boto3 client — no AWS keys in your code. Works with AWS, GCP, and Azure.

LLM Observability

hexr_llm() wraps any LLM client with OpenTelemetry. Per-agent cost attribution, token counts, and latency histograms — all automatic.

Agent-to-Agent Communication

Built-in A2A sidecar with JSON-RPC 2.0. Agents discover each other by name, exchange tasks, and stream results — all over mTLS.

​
Choose your deployment model

Hexr Cloud

Fully managed. Sign up and deploy in minutes. Pay-as-you-go with Hexr Compute Units (HCU).

Self-Hosted

Deploy on your own Kubernetes cluster with Terraform and Helm. Your infrastructure, your data.

​
Explore by topic

Architecture

Understand the 5-layer platform and how identity flows through every component.

Security

SPIFFE identity, OPA policies, zero-trust networking, and compliance frameworks.

Guides

Step-by-step tutorials for common agent patterns and use cases.