SDK

Python

The stillrunning-sdkpip package monitors any Python job, agent, or script. Decorate or wrap your code and StillRunning gets a ping on success or failure with the run's duration, plus optional cost, tokens, and model. Works with LangChain, CrewAI, agent scripts, Celery/RQ workers, and plain cron jobs. No runtime dependencies.

Install

terminal

pip install stillrunning-sdk

30-second quickstart

1. Create a workflow at stillrunning.ai/app/new and copy its token. 2. export STILLRUNNING_TOKEN=your_token_here. 3. Monitor your work:

job.py

from stillrunning_sdk import StillRunning

sr = StillRunning()  # reads STILLRUNNING_TOKEN

@sr.monitor
def nightly_job():
    ...

Every call is timed and reported, success on return, fail on exception (re-raised untouched).

Three ways to use it

Decorate a function:

decorator.py

@sr.monitor
def sync_customers():
    ...

Context manager, attach AI metrics inside the block:

context.py

with sr.track() as run:
    result = call_agent(prompt)
    run.meta(
        tokens_in=result.usage.input,
        tokens_out=result.usage.output,
        model="gpt-4o",        # model + tokens -> costUsd estimated automatically
    )

Run a callable inline:

run.py

answer = sr.run(
    lambda: call_agent(prompt),
    meta={"model": "claude-sonnet-4", "tokens_in": 1200},
)

Heartbeats and manual pings

manual.py

sr.heartbeat()                          # bare success ping
sr.ping(event="start", traceId="...")   # low-level: success | fail | start | log

Cost estimation

Give model plus token counts and cost is estimated from a built-in pricing table. Pass an explicit cost_usd for exact accounting, or extend the table:

cost.py

from stillrunning_sdk import register_model_pricing
register_model_pricing([(r"my-custom-model", (1.5, 6.0))])  # USD per 1M (input, output) tokens

Grouping multi-step runs

trace.py

from stillrunning_sdk import StillRunning, with_trace
sr = StillRunning()

with with_trace():
    sr.run(plan_step)
    sr.run(execute_step)   # both pings share one trace_id

Configuration

options.py

StillRunning(
    token=None,        # defaults to env STILLRUNNING_TOKEN
    base_url="https://stillrunning.ai",
    timeout=3.0,       # seconds; bounds each ping
    await_ping=True,   # False = fire-and-forget in a daemon thread
    on_error=None,     # callable(Exception) to observe ping failures
)

Monitoring never raises into your code: a failed ping routes to on_error and is otherwise swallowed.

Requirements

Python 3.9+. No runtime dependencies (standard library only).

You're set

Open your dashboard to watch runs land with duration, cost, and anomaly alerts.

Open dashboard