Skip to main content

AgentDB tells you when your agent stops behaving like itself. Causal lineage, time travel and behavioral baselines for any AI agent.

Project description

agentdb-sdk

Python SDK for AgentDB — the operational database for AI agents.

Install

pip install agentdb-sdk

Note: There is an unrelated package called agentdb on PyPI. Make sure to install agentdb-sdk (this package). The import is still from agentdb import AgentDB.

Quickstart

from agentdb import AgentDB

# Cloud (agentdb.zizka.ai)
db = AgentDB("agdb_live_xxxx")

# Self-hosted
db = AgentDB(host="http://localhost:8000")

Log an event

result = await db.log(
    agent="my-bot",
    event="tool_call",
    data={"tool": "search_web", "query": "competitor pricing"},
)
print(result.event_id)

Link events causally

# Log the user message
msg = await db.log(agent="my-bot", event="user_message",
                   data={"text": "why is my bill $200?"})

# Log the tool call that happened BECAUSE of the message
tool = await db.log(agent="my-bot", event="tool_call",
                    data={"tool": "get_billing"},
                    parent_id=msg.event_id)  # <-- causal link

# Log the response that happened BECAUSE of the tool call
await db.log(agent="my-bot", event="agent_response",
             data={"text": "I found an anomaly in your account"},
             parent_id=tool.event_id)

Why did something happen?

chain = await db.why(tool.event_id)
chain.print()
# user_message: {'text': 'why is my bill $200?'}  [14:32:01]
#     └── tool_call: {'tool': 'get_billing'}       [14:32:02]
#         └── agent_response: {'text': '...'}      [14:32:03]

Semantic search

results = await db.search("customer angry about billing")
for event in results:
    print(event.event, event.data)

Time travel

from datetime import datetime

state = await db.at("my-bot", datetime(2026, 5, 1, 15, 0))
print(state.state)

Query events

events = await db.query("my-bot", limit=100, event_type="tool_call")

Self-host

git clone https://github.com/Zizka-ai/agentdb
cd agentdb && cp .env.example .env
docker-compose -f infra/docker-compose.yml up

License

AGPL-3.0

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

agentdb_sdk-0.2.1.tar.gz (9.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

agentdb_sdk-0.2.1-py3-none-any.whl (9.7 kB view details)

Uploaded Python 3

File details

Details for the file agentdb_sdk-0.2.1.tar.gz.

File metadata

  • Download URL: agentdb_sdk-0.2.1.tar.gz
  • Upload date:
  • Size: 9.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.9.6

File hashes

Hashes for agentdb_sdk-0.2.1.tar.gz
Algorithm Hash digest
SHA256 1bdbdab0808f1e23c40fff8b1eeacf3c10e798cf6cca3040d21d329f950809d9
MD5 450672820f58334f4f9e66977352ae47
BLAKE2b-256 e6985730dbcc13df997a56626b5b62a328d8a120d224b62cb5b7863c618c40c2

See more details on using hashes here.

File details

Details for the file agentdb_sdk-0.2.1-py3-none-any.whl.

File metadata

  • Download URL: agentdb_sdk-0.2.1-py3-none-any.whl
  • Upload date:
  • Size: 9.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.9.6

File hashes

Hashes for agentdb_sdk-0.2.1-py3-none-any.whl
Algorithm Hash digest
SHA256 31eb638bef27164428954a46fb436d90e14cdbc780e190d213b595c5eb1241d0
MD5 7125f6b29d7191d17b8f8a879501f02a
BLAKE2b-256 3c436b184fadc6726fe25c4d3b97f49a9ba5d6fdeb65667f85e1572273d79970

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page