Stable

FastAPI Integration

Use decorators inside FastAPI services to capture runtime traces from API-triggered agent workflows and push sessions to Neurovn.

Install & Run

Install

Setup

pip install fastapi uvicornpip install neurovn

Run

Execute

NEUROVN_API_URL=https://agentic-flow.onrender.com uvicorn app:app --reload

Architecture Flow

Ingress

Request enters FastAPI route and starts workflow execution.

Trace

Decorated functions capture agent/tool steps and runtime metrics.

Persist

Completed session is published for canvas replay and estimate validation.

Implementation Snippets

Pattern

python

from fastapi import FastAPIfrom neurovn import traceapp = FastAPI()@trace.agent(name="API Agent", model="GPT-4.1", provider="OpenAI")def run_job(payload: dict) -> dict:    return {"ok": True}@app.post("/run")def run(payload: dict):    return run_job(payload)

Troubleshooting

Avoid blocking network calls in synchronous routes for predictable latency traces.

Mask sensitive request payload fields before custom logging.

Use one trace session per API request for deterministic auditability.

If you want a single canvas for each request lifecycle, open `trace.session(...)` at the route or service entrypoint instead of relying only on nested decorators.

Related Integrations

code to-canvas

Integration guide

decorator

Integration guide

cli

Integration guide

Backend contracts: `/api/estimate`, `/api/traces/sessions`, `/api/canvases`