• Causal, not just traces. Each event carries a parent_id. db.why(event_id) walks back to the user message, wrong tool, or bad context that started it.
• Time-travel. db.at(agent, timestamp) rebuilds exactly what the agent knew at any past moment.
• Tamper-evident audit trail. Every decision is logged with a checksum, so you have a verifiable record for EU AI Act Article 12.
• Drift detection. See when an agent's behavior shifts from its baseline.
• Self-host or cloud. Run it on your own Postgres with one Docker command, or use ZizkaDB Cloud. AGPL-3.0, no per-trace billing; self-hosted data never leaves your infrastructure.

Contents

• Quickstart (60 seconds)
• Integrations
• Connect your agent
• What it does
• Audit trail and EU AI Act Article 12
• How it works
• Use with your AI assistant (MCP)
• ZizkaDB vs. tracing tools
• Cloud, FAQ and docs
• Contributors

Quickstart (60 seconds)

From zero to your first causal chain with one command. No repo clone needed.

1. Start Docker. Docker must be running. The first image pull can take 5–10 minutes; later starts take seconds.

2. Install and run. This downloads config and pre-built images, starts Postgres, Qdrant, Redis, the API and the dashboard, then runs a demo agent:

curl -fsSL https://raw.githubusercontent.com/Zizka-ai/ZizkaDB/main/scripts/quickstart-remote.sh | bash

3. See why. The demo prints the causal chain behind the agent's tool call:

tool_call · lookup_order · ORD-8842
  └── llm_response · gpt-4o
        └── user_message · Why was my order delayed?

4. Open the dashboard. localhost:3001 → Open my dashboard → Activity → click any event → Why? (causal) tab.

The self-hosted dashboard is just your dashboard: no signup, no email, no account. Accounts and plans exist only on ZizkaDB Cloud.

Run the demo again anytime: pip install zizkadb-sdk && zizkadb demo

Self-host from a clone

git clone https://github.com/Zizka-ai/ZizkaDB.git && cd ZizkaDB
bash scripts/setup-local.sh

| Service | URL | |---------|-----| | API | http://localhost:8000 | | Dashboard | http://localhost:3001 → Open my dashboard | | Swagger | http://localhost:8000/swagger |

Self-host on a server

On a machine other people can reach, set these in infra/.env before starting the stack:

ENV=production                      # turns off the one-click button and dev API keys
DEV_API_KEY=<random>                # must not be the default, or the API refuses to start
JWT_SECRET=<random>                 # openssl rand -hex 32 (also JWT_REFRESH_SECRET)
DEPLOYMENT_MODE=self_hosted
SELFHOST_ADMIN_TOKEN=<long-random>  # python -c "import secrets; print(secrets.token_urlsafe(32))"

The dashboard then asks for the admin token instead of showing the one-click button. Without SELFHOST_ADMIN_TOKEN, dashboard login stays disabled. Everyone who has the token signs in to the same single owner workspace. Check your config with bash scripts/validate-selfhost-config.sh --production. Full steps: wiki/Self-Hosting.

Full guide: DEVELOPMENT.md · Troubleshooting: wiki/Troubleshooting.md

Integrations

Works with the stack you already use — add one package and every step is logged with its cause.

Integration	Install	What you get
Python	pip install zizkadb-sdk	Any Python agent
TypeScript	npm install zizkadb-sdk	Any JavaScript / TypeScript agent
LangChain	pip install zizkadb-langchain	Drop-in callback handler
CrewAI	pip install zizkadb-crewai	Crew logger for your agents
LiveKit	pip install zizkadb-livekit	Voice agents — one call, one session
MCP	uvx zizkadb-mcp	Ask Cursor or Claude why
REST API	Swagger docs	Any language

Click a name for its setup guide. New project? Scaffold one with zizkadb init my-agent --template basic.

Connect your agent

import asyncio
from zizkadb import ZizkaDB
async def main():
    async with ZizkaDB(host="http://localhost:8000") as db:
        user = await db.log(agent="my-bot", event="user_message", data={"text": "Why is my order late?"})
        tool = await db.log(agent="my-bot", event="tool_call", data={"tool": "lookup_order"}, parent_id=user.event_id)
        (await db.why(tool.event_id)).print()
asyncio.run(main())

import { ZizkaDB } from 'zizkadb-sdk'
const db = new ZizkaDB({ host: 'http://localhost:8000' })
const user = await db.log({ agent: 'my-bot', event: 'user_message', data: { text: 'Why is my order late?' } })
const tool = await db.log({ agent: 'my-bot', event: 'tool_call', data: { tool: 'lookup_order' }, parentId: user.eventId })
;(await db.why(tool.eventId)).print()

From the terminal: zizkadb why . Full guide: CONNECT.md

What it does

| Function | What you get | | --- | --- | | db.why(event_id) | The causal chain behind any event | | db.at(agent, timestamp) | What the agent knew at a past moment | | db.search(query) | Semantic search over the agent's history | | db.context_for(agent, task) | Relevant past events, ready to inject into the next prompt | | db.baseline(agent) | Drift detection: alerts when agent behavior shifts from past sessions | | db.forget(key, value) | GDPR erasure by metadata filter, including the search index |

Audit trail and EU AI Act Article 12

Article 12 of the EU AI Act requires high-risk AI systems to automatically keep logs of what they did. ZizkaDB gives you that record:

• Checksum-backed decision logs. Every event is stored with a SHA-256 checksum of its content, so any later edit is detectable.
• Causally linked history. Each decision points to the event that caused it, so an auditor can follow the full chain.
• Session replay. Step through any past session event by event.
• Time-travel debugging. Rebuild exactly what the agent knew at any moment with db.at().
• Drift detection. db.baseline() flags when an agent starts behaving differently from its history.

How it works

flowchart LR
    A[Your agent<br/>SDK · LangChain · CrewAI · LiveKit] -->|events + parent_id| B[ZizkaDB API]
    D[Dashboard] --> B
    M[MCP server<br/>Cursor · Claude] --> B
    B --> P[(PostgreSQL<br/>source of truth)]
    B --> Q[(Qdrant<br/>semantic search)]
    B --> R[(Redis<br/>cache)]

• Causal lineage lives in Postgres: every event stores its parent, and why() walks the chain with a recursive query. No separate graph store.
• Every event is written twice: to Postgres for structured queries and to Qdrant for semantic search. Design decisions: docs/adr/.

Use with your AI assistant (MCP)

Ask Cursor or Claude "why did support-bot call lookup_order?" and get the chain back. Add this to your MCP config:

{
  "mcpServers": {
    "zizkadb": {
      "command": "uvx",
      "args": ["zizkadb-mcp"],
      "env": { "ZIZKADB_HOST": "http://localhost:8000" }
    }
  }
}

For ZizkaDB Cloud, use ZIZKADB_API_KEY instead. Setup for each client: mcp/README.md. The MCP server is MIT-licensed.

ZizkaDB vs. tracing tools

Tools like Langfuse and LangSmith observe span trees. ZizkaDB audits decisions.

| | ZizkaDB | Typical LLM tracing tools | | --- | :---: | :---: | | Explicit cause → effect links | ✅ | Span nesting | | One-call root cause (db.why()) | ✅ | Manual trace reading | | Time-travel to past agent state | ✅ | — | | Memory for future runs (db.context_for()) | ✅ | — | | Pricing | Free self-host (AGPL) | Often per-trace |

Contributors

Thanks to everyone who has helped build ZizkaDB. Want to join? Read CONTRIBUTING.md or pick up an open issue.

saadamjad	Zizka-ai	arshadgit23	saadwashmen
Subhajitdas99	mikeaig4real	aqilaziz	lamenting-hawthorn
abdurrehman616	HafizHamzaShahid	eaz1337	mbilalzeeshan
AbdelazizBs	Mephistopheles9631	Qalbeabbas-12	towfiq-ul

_{AGPL-3.0 · MCP server MIT · Disable telemetry: export ZIZKADB_TELEMETRY=false
This repo is the open-source self-host stack. The operator console and VPC deploy live in a private repo (why).}