Quickstart

Call Korely from your own code, any language over REST, or Python and Node via pip install korely-memory / npm install korely-memory. The path for building persistent memory into an app, an agent, or a backend.

REST API & Python SDK

1. Install

pip install korely-memory

npm install korely-memory

# No install needed, the REST API works with any HTTP client.

The Python and Node SDKs are thin, zero-dependency clients over the REST API, same package name (korely-memory) on PyPI and npm. Every method calls the REST API; the Python SDK's project and key methods call the admin API, which takes an admin key.

2. Get an API key

The fastest way is one command. pip install korely-memory ships the korely CLI; korely init --agent mints a free Hobby key (kor_live_…) and saves it locally, so the CLI works immediately. Any language can mint one over REST instead:

korely init --agent

curl -X POST https://api.korely.ai/v1/agents/init \
  -H "Content-Type: application/json" \
  -d '{"agent_caller": "my-app"}'

The REST answer is 201 Created with the key (shown once), "region": "eu-hel1", the label of the Helsinki server that stores your data, and "processing_region": "global", where the model that reads your memories into facts runs (Google Gemini). Send "processing_region": "eu" in the body, or run korely init --agent --processing-region eu, to have it run on Scaleway in Paris instead; your data is stored in the EU either way.

See Sign up as an agent for the full flow. To call Korely from the Python or Node SDK (rather than the CLI), set the key as an environment variable, the SDK reads it automatically:

export KORELY_API_KEY=kor_live_...

An account made with korely init --agent is the key: it has no login. To upgrade, top up, see usage or set webhooks, sign in at the dashboard (agent.korely.ai) and bring the key in on the Account page: everything it wrote moves into that account, and the key keeps working (details).

3. Your first memory

from korely_memory import Korely

korely = Korely()  # reads KORELY_API_KEY from the environment

# Remember something your agent learned about an end user
korely.add(
    "Maria prefers email over Slack, and her renewal is in October.",
    user_id="customer-4812",
)

# Later, even in a new session, pull a prompt-ready block back
ctx = korely.get_context(
    query="how does Maria like to be contacted?",
    user_id="customer-4812",
)
print(ctx.context)  # drop straight into your system prompt

import { Korely } from "korely-memory";

const korely = new Korely(); // reads KORELY_API_KEY from the environment

// Remember something your agent learned about an end user
await korely.add(
  "Maria prefers email over Slack, and her renewal is in October.",
  { user_id: "customer-4812" },
);

// Later, even in a new session, pull a prompt-ready block back
const ctx = await korely.getContext({
  query: "how does Maria like to be contacted?",
  user_id: "customer-4812",
});
console.log(ctx.context); // drop straight into your system prompt

# Remember something
curl -X POST https://api.korely.ai/v1/memories \
  -H "Authorization: Bearer kor_live_..." \
  -H "Content-Type: application/json" \
  -d '{"content": "Maria prefers email over Slack, and her renewal is in October.", "user_id": "customer-4812"}'

# Later, pull a prompt-ready context block back
curl "https://api.korely.ai/v1/context?query=how+does+Maria+like+to+be+contacted%3F&user_id=customer-4812" \
  -H "Authorization: Bearer kor_live_..."

That's the whole loop: add to remember, get_context (or search) to recall. Behind add, Korely extracts typed, bi-temporal facts and builds a graph, you just hand it text.

4. Search, and see what comes back

get_context is the recall path you'll reach for most, it assembles the active typed facts most relevant to the query, plus the most relevant memories, into a block you drop straight into a prompt. When you'd rather inspect raw ranked hits, search runs a semantic vector search and returns memories ordered by relevance score:

hits = korely.search("contact preference", user_id="customer-4812")
for h in hits:
    print(h.score, h.snippet)

curl -X POST https://api.korely.ai/v1/memories/search \
  -H "Authorization: Bearer kor_live_..." \
  -H "Content-Type: application/json" \
  -d '{"query": "contact preference", "user_id": "customer-4812"}'

{
  "results": [
    {
      "id": "mem_8f2c1a",
      "score": 0.91,
      "snippet": "Maria prefers email over Slack, and her renewal is in October.",
      "user_id": "customer-4812",
      "agent_id": null,
      "metadata": {}
    }
  ]
}

5. What you can do

The Python and Node SDKs have the same memory, fact and account methods (camelCase in Node), each mapping to a REST endpoint, from add to ping, audit and delete_account; the Python SDK also has the admin-key methods for projects and keys. The headlines (the full list is in the SDK reference):

Full detail with examples: the SDK reference and the REST API reference. Limits per plan are on the pricing page.

Next steps

You're up and running. Where to go from here:

• Explore the full SDK surface: the SDK reference covers all methods with examples for Python and Node.
• Use the REST API directly: the API reference documents every endpoint, request body, and response shape.
• Understand how memory works: read the memory model to learn how Korely extracts facts, builds a graph, and handles contradictions over time.
• Wire it into a framework: the integrations index has step-by-step guides for the Vercel AI SDK, LangGraph, n8n, the OpenAI Agents SDK, CrewAI and LlamaIndex; the OpenClaw plugin is not published yet.
• Use the CLI: the CLI reference covers korely add, korely search, korely context, and the other subcommands, useful for scripts and one-off queries.