OpenAI and OpenAI-compatible

​

Install

pip install sakshi-sdk openai

​

Wrap the client

from openai import OpenAI

client = OpenAI()

reply = client.chat.completions.create(
    model="gpt-5.2",
    messages=[{"role": "user", "content": "Assess this applicant."}],
)

from openai import OpenAI
from sakshi.middleware import watch_openai_compatible

client = watch_openai_compatible(OpenAI())

with sakshi.witness(agent_id, client_ref="APP-2026-04471") as decision:
    reply = client.chat.completions.create(
        model="gpt-5.2",
        messages=[{"role": "user", "content": "Assess this applicant."}],
    )
    decision.action(outcome="approved", mode="auto")

​

Self-hosted and sovereign runtimes

from openai import OpenAI
from sakshi.middleware import watch_openai_compatible

# An Ollama server on the same host, or a vLLM service in the cluster.
local = watch_openai_compatible(
    OpenAI(base_url="http://ollama:11434/v1", api_key="ollama")
)

with sakshi.witness(agent_id) as decision:
    reply = local.chat.completions.create(
        model="qwen2.5:1.5b",
        messages=[{"role": "user", "content": "Assess this applicant."}],
    )

​

What gets recorded

• The provider, derived from the base URL as above.
• model_requested versus model_served, so a silent model swap is visible.
• system_fingerprint when the provider returns one.
• Prompt and completion token usage.
• The finish reason and the call latency.
• With content capture on, the last message you sent and the response text, each truncated to content_limit.

​

Capture options

watch_openai_compatible(client, capture_content=True, content_limit=1000)

• capture_content defaults to True. Content capture is safe by design: the platform tokenizes PII at ingest, before storage or hashing. Set it to False for a minimal-capture deployment where only metadata, tokens, and identity are recorded.
• content_limit caps how much of each message and response is stored. The default is 1000 characters.