Smello

Capture outgoing and incoming HTTP requests, pytest results, Python logs, and unhandled exceptions in a local web dashboard.

Like Mailpit, but for your entire debug output.

Setup

Install the client SDK and the server:

pip install smello smello-server

Start the server:

smello-server

Run your code with Smello:

smello run my_app.py
smello run pytest tests/
smello run uvicorn app:app

That's it. Smello activates before your code runs and captures outgoing traffic, pytest results, unhandled exceptions, and optional Python log records. No code changes needed.

Subprocess instrumentation propagates automatically through PYTHONPATH, so smello run gunicorn app:app also captures traffic from worker processes.

CLI flags map 1:1 to the SMELLO_* env vars: --server, --capture-host, --ignore-host, --capture-all / --no-capture-all, --redact-header, --redact-query-param, --capture-tests / --no-capture-tests, --capture-logs, --log-level, --app, --session. The action-only --clear flag does not set an environment variable.

Clear all events with smello clear, or begin a command with an empty timeline:

smello run --clear my_app.py

Tag events with --app and --session to isolate a debugging run:

smello run --app myapp --session debug-payment python scripts/checkout.py

Using smello.init() instead

If you prefer to activate Smello from within your code (e.g., for programmatic configuration or projects with a custom sitecustomize.py):

import smello
smello.init()  # activates only when SMELLO_URL is set

Framework middleware

To capture incoming requests, add the Smello middleware to your web framework:

FastAPI:

from smello.integrations.fastapi import SmelloMiddleware
from fastapi import FastAPI

app = FastAPI()
app.add_middleware(SmelloMiddleware, ignore_paths=["/health"])

Django:

# settings.py
MIDDLEWARE = [
    "smello.integrations.django.SmelloMiddleware",
    ...
]
SMELLO_IGNORE_PATHS = ["/health/", "/admin/"]

Then run with smello run:

smello run uvicorn app:app        # FastAPI
smello run manage.py runserver    # Django

The middleware captures method, path, status code, duration, route pattern, client IP, and request/response bodies. Unhandled exceptions are captured with full tracebacks. When Smello is inactive (no server URL configured), the middleware passes requests through without capturing anything.

Google Cloud libraries

Many Google Cloud Python libraries — BigQuery, Firestore, Pub/Sub, Analytics Data API (GA4), Vertex AI, Speech-to-Text, Vision, Translation, and others — use gRPC under the hood. Smello captures these calls automatically:

smello run my_bigquery_script.py

Any library that calls grpc.secure_channel() or grpc.insecure_channel() is automatically captured.

What Smello captures

Outgoing HTTP requests — method, URL, headers, body, response status/headers/body, duration, and library used (requests, httpx, aiohttp, grpc, or botocore).

Failed outgoing calls retain exception details. If a library raises after receiving a response, Smello stores both the response status and headers and the exception.

Incoming HTTP requests (via FastAPI or Django middleware) — method, path, route pattern, status code, duration, client IP, request/response headers and bodies, plus exception tracebacks if a handler raises.

Pytest executions (enabled by default): Smello supports pytest 7 and later and records one event per test function or method invocation. Parametrized cases remain separate. Install the integration with pip install "smello[pytest]" if your project doesn't already include pytest.

Each event includes the outcome, fixture names, source location, total and per-phase timings, and failure details. It also carries runtime trace and span IDs plus stable directory, file, class, and unparameterized case memberships. Outgoing Requests, HTTPX, aiohttp, botocore, and gRPC calls inherit those memberships and appear as child operations. The dashboard always shows runtime parenthood and can add one virtual grouping by an available pytest level or remote host without changing event deep links. Virtual groups merge only across adjacent siblings, which preserves record order. Timeline filters retain matching records' runtime ancestors, so filtered children remain connected to their runtime context. The traceback generated by pytest may include repr() output for parameter, fixture, or local variable values.

See Debug pytest tests with Smello for a runnable example and dashboard walkthrough.

Unhandled exceptions (enabled by default) — exception type, message, full traceback, and stack frames with source context.

Log records (opt-in via capture_logs=True) — level, logger name, message, source location, and extra attributes.

Smello redacts sensitive headers (Authorization, X-Api-Key) by default and optionally redacts query string parameters.

Configuration

smello.init(
    server_url="http://localhost:5110",       # where to send captured data

    # HTTP capture
    capture_hosts=["api.stripe.com"],         # only capture these hosts
    capture_all=True,                          # capture everything (default)
    ignore_hosts=["localhost"],               # skip these hosts
    redact_headers=["Authorization"],         # replace header values with [REDACTED]
    redact_query_params=["api_key", "token"], # replace query param values with [REDACTED]

    # Tests, logs, and exceptions
    capture_tests=True,                        # capture pytest executions (default)
    capture_exceptions=True,                   # capture unhandled exceptions (default)
    capture_logs=False,                        # capture log records (opt-in)
    log_level=30,                              # minimum log level to capture (WARNING)
    ignore_loggers=["uvicorn.access"],         # suppress noisy framework loggers

    # Tagging
    app="myapp",                               # tag events with an application name
    session="debug-payment",                   # tag events with a session ID
)

All parameters fall back to SMELLO_* environment variables when not passed explicitly:

The server URL is the activation signal — init() does nothing unless server_url is passed or SMELLO_URL is set. Boolean env vars accept true/1/yes and false/0/no (case-insensitive). List env vars are comma-separated.

Smello Server deletes events older than seven days when it starts and every hour after that. Set SMELLO_RETENTION_DAYS on the server process to change the retention period, or set it to 0 to keep events indefinitely. The server rejects negative or malformed values at startup.

Calls to the OpenAI, Anthropic, and Gemini APIs render as a readable conversation in the dashboard, with the system prompt, tool calls, and token usage broken out. The raw JSON stays one tab away.

Supported libraries

• requests — patches Session.send()
• httpx — patches Client.send() and AsyncClient.send()
• aiohttp — injects TraceConfig lifecycle hooks to capture async HTTP traffic
• grpc — patches insecure_channel() and secure_channel() to intercept unary-unary calls
• botocore — patches URLLib3Session.send() to capture boto3 / AWS SDK traffic
• pytest: loads a plugin that captures each test function or method invocation

Requires

• Python >= 3.10
• smello-server running locally

Links

• Documentation
• Source & Issues
• Changelog
• smello-server on PyPI