Operational Database For AI Agents

Fix failures in minutes, not after customer complaints.

Works with your stack

Python SDK

pip install zizkadb-sdk

TypeScript SDK

npm i zizkadb-sdk

Cursor MCP

Native tools

REST API

Any language

Open source

AGPL · self-host

Managed cloud

db.zizka.ai

How it works

Three ways to connect

Managed cloud, MCP in Claude / Cursor / OpenAI, or self-hosted — same SDKs and API.

Managed cloud

We run the stack at db.zizka.ai. Sign up, create an agent, copy your API key.

Create account

Create agent & key

Dashboard → create agent → copy zizkadb_live_…

pythonpip install zizkadb-sdk
from zizkadb import ZizkaDB
db = ZizkaDB("zizkadb_live_xxxx")
await db.log(agent="my-bot", event="started", data={})

MCP

Plug into Claude Desktop, Cursor, or OpenAI MCP. No app refactor.

Install

pip install zizkadb-mcp or uvx zizkadb-mcp

Paste & reload

Add to Cursor mcp.json or Claude Desktop config.

json{
  "mcpServers": {
    "zizkadb": {
      "command": "uvx",
      "args": ["zizkadb-mcp"],
      "env": { "ZIZKADB_API_KEY": "zizkadb_live_xxxx" }
    }
  }
}

Self-hosted

Run on your laptop or VPS — AGPL, free forever. Point SDKs at your API URL.

Clone & start

GitHub → bash scripts/setup-local.sh

SDK

pip install zizkadb-sdk · npm i zizkadb-sdk

pythonfrom zizkadb import ZizkaDB
db = ZizkaDB(host="http://localhost:8000")
await db.log(agent="my-bot", event="started", data={})

GitHub → · Docs

See what changes when you can replay any session

Same incident. Without ZizkaDB you guess. With it you see the exact decision trail.

WITHOUT ZIZKADB

Your team today

Customer

Your bot told me refunds take 30 days. That is wrong.

Your team

Can you send a screenshot? We will check the logs.

Engineering

Logs show nothing useful. Maybe a prompt issue?

Hours lost. Still no root cause.

WITH ZIZKADB

Same incident, 2 minutes later

Drift alert

Refund answers changed after prompt v2 deploy.

Replay

Agent skipped policy doc, used outdated FAQ chunk.

Fix

Roll back prompt. Confirm baseline restored.

Root cause found. Prompt rolled back.

For engineering teams

Ship agents faster. Debug them in production.

Whether you are building a single agent or a multi-agent system, ZizkaDB fits how dev teams work.

⚙️

Integrate in an afternoon

Log from Python, TypeScript, MCP, or REST. Self-host with Docker or use managed cloud. One API key per agent. No heavy infra project.

5-minute setup

🔍

Debug production incidents

Prompt change broke behavior? Tool call ignored? Replay the full session, trace the decision chain, and find root cause without digging through logs.

Start free trial

Trust & security

Production-ready from day one

📦

Open source

AGPL license. Inspect, fork, self-host free.

🏠

Self-host free

Docker Compose on your own infrastructure.

🔒

Checksum-backed events

Tamper-evident decision history.

🇪🇺

Managed cloud

db.zizka.ai when you want zero ops.

Pricing

Start free. Scale when you need to.

Full monitoring and session replay on every plan.

Self-Hosted

Freeforever

✓ Full feature set
✓ Your infrastructure
✓ Docker Compose
✓ Community support

Setup guide

POPULAR

Pro

€39/ month

✓ 100M events
✓ 90-day retention
✓ 3 projects
✓ Email support

Start free trial

No credit card needed

Team

€99/ month

✓ Up to 1B events/mo
✓ 1-year retention
✓ 10 seats
✓ Priority support

Start free trial

No credit card needed

Compare

Operational database vs traces, memory, and vectors

Vector DBs store embeddings. Traces show spans. ZizkaDB stores agent decisions, lineage, and drift.

Capability	LangSmith	Mem0	Pinecone	ZizkaDB
Agent event logging	✓	✗	✗	✓
Causal lineage	~	✗	✗	✓
Time travel (state at T)	✗	✗	✗	✓
Semantic search on history	✗	✓	✓	✓
Behavioral baseline / drift	✗	✗	✗	✓
Cross-agent fleet queries	✗	✗	✗	✓
Self-host free	✓	✓	✗	✓

~ = partial support. Verify competitor docs before external debates.

View on GitHub →Full comparison

Fix before production breaks

Start free on db.zizka.ai. Self-host anytime.

Start free trial Read the docs GitHub →