Skip to main content
Pre-release

This release is a pre-release and may not be stable for production use.

KAgent OpenAI Agents SDK Integration

OpenAI Agents SDK integration for KAgent with A2A (Agent-to-Agent) protocol support, session management, and optional skills integration.


Quick Start

from kagent.openai import KAgentApp
from agents.agent import Agent

# Create your OpenAI agent
agent = Agent(
    name="Assistant",
    instructions="You are a helpful assistant.",
    tools=[my_tool],  # Optional
)

# Create KAgent app
app = KAgentApp(
    agent=agent,
    agent_card={
        "name": "my-openai-agent",
        "description": "My OpenAI agent",
        "version": "0.1.0",
        "capabilities": {"streaming": True},
        "defaultInputModes": ["text"],
        "defaultOutputModes": ["text"],
    },
    kagent_url="http://localhost:8083",
    app_name="my-agent",
)

# Run
fastapi_app = app.build()
# uvicorn run_me:fastapi_app

Agent with Skills

Skills provide domain expertise through filesystem-based instruction files and helper tools (read/write/edit files, bash execution). We provide a function to load all skill-related tools. Otherwise, you can select the ones you need by importing from kagent.openai.tools.

from agents.agent import Agent
from kagent.openai import get_skill_tools

tools = [my_custom_tool]
tools.extend(get_skill_tools("./skills"))

agent = Agent(
    name="SkillfulAgent",
    instructions="Use skills and tools when appropriate.",
    tools=tools,
)

See skills README for skill format and structure.


Session Management

Sessions persist conversation history in KAgent backend:

from agents.agent import Agent
from agents.run import Runner
from kagent.openai._session_service import KAgentSession
import httpx

client = httpx.AsyncClient(base_url="http://localhost:8083")
session = KAgentSession(
    session_id="conversation_123",
    client=client,
    app_name="my-agent",
)

agent = Agent(name="Assistant", instructions="Be helpful")
result = await Runner.run(agent, "Hello!", session=session)

Local Development

Test without KAgent backend using in-memory mode:

app = KAgentApp(agent=agent, agent_card=agent_card, kagent_url="http://localhost:8083", app_name="test-agent")

fastapi_app = app.build_local()  # In-memory, no persistence

Deployment

Standard Docker deployment:

FROM python:3.13-slim
WORKDIR /app
COPY requirements.txt .
RUN pip install -r requirements.txt
COPY agent.py .
CMD ["uvicorn", "agent:fastapi_app", "--host", "0.0.0.0", "--port", "8000"]

Load /config/config.json with KAgentConfig.from_file() to connect to kagent.


Architecture

Component Purpose
KAgentApp FastAPI application builder with A2A support
KAgentSession Session persistence via KAgent REST API
OpenAIAgentExecutor Executes agents with event streaming

Environment Variables


Examples

See samples/openai/ for complete examples:

  • basic_agent/ - Simple agent with custom tools
  • More examples coming soon

See Also


License

See repository LICENSE file.

Runtime configuration file

Managed runtimes receive /config/config.json from the controller. For a standalone process, create that file and load it with KAgentConfig.from_file():

{"runtime":{"control_plane_url":"http://localhost:8083","gateway_url":"http://localhost:8083","name":"my-agent","namespace":"default","log_level":"info"}}

Provider credentials and standard OTEL_* overrides remain environment inputs.

Metadata

Release files for kagent-openai 1.0.0a15

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for kagent-openai 1.0.0a15
File Size Uploaded Provenance
kagent_openai-1.0.0a15.tar.gz 15.7 kB Without provenance Details

Built distribution (wheel)

Table of built distributions (wheels) for kagent-openai 1.0.0a15
File Interpreter ABI Platform
kagent_openai-1.0.0a15-py3-none-any.whl Python 3 none any Details

Total release size: 29.8 kB

Release files / kagent_openai-1.0.0a15.tar.gz

Download URL kagent_openai-1.0.0a15.tar.gz
Size 15.7 kB
Tags Source
SHA-256 checksum
How to use checksums
7d3993c3369eddc9276cde4f715cd83ecc1fbfd78bc064a17d7e4f2033e077be
BLAKE2b-256 checksum
How to use checksums
545020b1ee001c0c0b88e6e2ff6d3f9f972927f3cf2f4760c84ce9e921f22cab
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.13.1 {"installer":{"name":"uv","version":"0.13.1","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}

Release files / kagent_openai-1.0.0a15-py3-none-any.whl

Download URL kagent_openai-1.0.0a15-py3-none-any.whl
Size 14.1 kB
Tags Python 3
SHA-256 checksum
How to use checksums
6f2b2c99c5f62f8984e90796e77721cf07b84362f46c720bd9bf06b139cd8d41
BLAKE2b-256 checksum
How to use checksums
949fc723d7eb6b95566d3f8d7cd539b207fec9d468f24aa61506971237092793
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via uv/0.13.1 {"installer":{"name":"uv","version":"0.13.1","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}

Security for kagent-openai 1.0.0a15

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

No provenance found

None of the files in this release were uploaded with provenance metadata.

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

Release history Release notifications | RSS feed

This release
Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page