ppx-openai

OpenAI Assistants API integration for the Preference Profile Exchange (PPX).

Install

pip install ppx-openai

Requires Python 3.11+. Published version: 0.1.0a1 (PyPI).

Alpha, tracking a draft specification. Expect breaking changes. The version is 0.1.0a1 on PyPI and 0.1.0-alpha.1 on npm — the same release in each ecosystem's required format.

Assistants API — minimal example

from openai import OpenAI
from ppx_client import PpxClient
from ppx_openai import ppx_tool_schema, handle_run

oai = OpenAI()
ppx = PpxClient("https://api.provider.app")

assistant = oai.beta.assistants.create(
    name="scent-advisor",
    model="gpt-4o-mini",
    instructions=(
        "Recommend fragrances. Use lookup_user_preference before suggesting "
        "anything opinionated — the user's PPX profile has their preferences."
    ),
    tools=[ppx_tool_schema()],
)

thread = oai.beta.threads.create()
oai.beta.threads.messages.create(
    thread_id=thread.id, role="user",
    content="Suggest something for a hot, humid evening.",
)
run = oai.beta.threads.runs.create(thread_id=thread.id, assistant_id=assistant.id)

# Handler loops over requires_action → tool output → continue
final = handle_run(
    thread_id=thread.id,
    run=run,
    openai_client=oai,
    ppx_client=ppx,
    grant_token=my_grant_token,
)

messages = oai.beta.threads.messages.list(thread_id=thread.id)
print(messages.data[0].content[0].text.value)

Every tool call routes through the same grant-scoped read path as the browser flow. The model can never see claims the user didn't authorize.

License

Apache-2.0.