Onboarding: Anthropic

Add persistent memory to your Anthropic Claude-powered application. Works with the Anthropic Python and TypeScript SDKs.

Best path: Gateway (change one URL) or SDK (one import) Time: 2 minutes Reliability: ~99% (Gateway) / ~95% (SDK)

Option A: Gateway (Recommended)

The Gateway speaks the OpenAI chat completions protocol. To use it with Anthropic models, you point the Gateway at Anthropic's API.

Step 1: Get an API key

Step 2: Use the OpenAI SDK pointed at Anthropic

from openai import OpenAI

client = OpenAI(
    base_url="https://api.hippocortex.dev/v1",
    api_key="hx_live_...",
    default_headers={
        "X-LLM-API-Key": "sk-ant-...",
        "X-LLM-Base-URL": "https://api.anthropic.com",
    },
)

response = client.chat.completions.create(
    model="claude-sonnet-4-20250514",
    messages=[{"role": "user", "content": "Hello!"}],
)

Step 3: Verify

Check dashboard.hippocortex.dev to see captured events.

Option B: SDK

The Hippocortex SDK supports the native Anthropic SDK through auto-instrumentation.

Step 1: Install

pip install hippocortex==1.2.1
# or
npm install @hippocortex/sdk@1.2.1

Step 2: Set your API key

export HIPPOCORTEX_API_KEY=hx_live_...

Step 3: Add one import

Python:

import hippocortex.auto
from anthropic import Anthropic

client = Anthropic()
# Every call now has memory.

TypeScript:

import '@hippocortex/sdk/auto'
import Anthropic from '@anthropic-ai/sdk'

const anthropic = new Anthropic()
// Every call now has memory.

What to expect

After a few interactions, your agent starts receiving synthesized context from past sessions. The compilation pipeline runs automatically in the background.