Catalog PyPI
PACKAGE EXPLORERView in registry

aicery-sdk

PyPI

Python client for the Aicery AI agent runtime — runs, agents, and SSE streaming

agentsailanggraphllmorchestrationruntimetrace
Versions collected1
Direct dependencies2From release 0.0.2
Last observed21 Oct, 10:07 UTC

README

Version 0.0.2 documentation

aicery-sdk

Official Python client for the Aicery AI agent runtime.

About Aicery

Aicery is an open-source AI agent runtime — orchestration, tools, trace, and replay. It is the engine under your application, not a vertical SaaS product.

Agents are replaceable. Runtime is permanent.

The runtime runs LangGraph agents, executes tools safely, records step-by-step traces, and replays runs with frozen mocks. You bring UI, business rules, and data connectors; Aicery handles runs, providers, and observability.

This package (aicery-sdk) is a thin HTTP client for the runtime API (/v1/runs, /v1/agents, SSE streaming). It does not include the runtime server — run the Aicery stack locally or point the client at your deployment.

Install

pip install aicery-sdk

Requires Python 3.11+.

Prerequisites

A running Aicery runtime (default: http://localhost:8000). From the main repo:

git clone https://github.com/mrkkyatilla/aicery.git
cd aicery
make install && make up && make migrate

Default API key for local dev: dev.

Quick start

from aicery_sdk import AiceryClient

client = AiceryClient(base_url="http://localhost:8000", api_key="dev")

run = client.create_run(agent_id="echo", input="hello", execute=True)
final = client.get_run(run.id)

print(final.status)       # completed
print(final.output_text)  # echo:hello

create_run may return status: pending immediately; call get_run for the final output.

Configuration

From aicery.yaml (recommended in projects):

client = AiceryClient.from_config("aicery.yaml")

From environment:

client = AiceryClient.from_env()
# API_KEY or AICERY_API_KEY (default: dev)
# AICERY_RUNTIME_URL (default: http://localhost:8000)

List agents

agents = client.list_agents()
for agent in agents:
    print(agent["id"], agent.get("name"))

Streaming

run = client.create_run(agent_id="research", input="Summarize README", execute=True)

for event in client.stream_run(run.id):
    if event.event == "token":
        print(event.data, end="", flush=True)

API surface (v0.0.x)

MethodDescription
create_run(...)Start or execute an agent run
get_run(run_id)Fetch run status and output
list_agents()List available agents
stream_run(run_id)SSE token/step stream
resume_run(...)HITL resume after approval
from_config(path)Load client from aicery.yaml
from_env()Load client from environment

Development

From the monorepo (editable install):

pip install -e ./sdk/python
pytest sdk/python/tests

Downloads & popularity

Awaiting observations

No metric observations yet

Download counts and popularity will appear here as the collector receives them.

Versions 1

Collected releases
VersionPublishedDirect dependenciesOpen version
0.0.210 Jun 20262

Package advisories

0
No matching advisories collected

This describes our collected data; it is not a security assessment.

Repository

PENDING
mrkkyatilla/aicery

Shared repository data · GitHub

License detected in repository
Not collected

Metadata history

Meaningful changes
Current collected state Current

Observed since 21 Oct, 10:07 UTC

No superseded metadata states collected. Unchanged observations do not create new revisions.

Record identity

package:7abd3ee16b567698676a5e26ce440a8e97147e6b8ad263b1b7609e4351fe1429