---
title: "Python | Speedscale Docs"
description: "Python guidance for Speedscale and proxymock, including proxy setup, TLS trust, demo app details, and a first-success workflow."
canonical: "https://docs.speedscale.com/reference/languages/python/"
---

[Documentation index](https://docs.speedscale.com/reference/llms.txt) | [Markdown](https://docs.speedscale.com/reference/languages/python/index.md) | [Canonical page](https://docs.speedscale.com/reference/languages/python/)

# Python

Python is fully supported by Speedscale. Use this page for Python-specific proxy settings, TLS trust configuration, demo guidance, and the proxymock local workflow.

*   Support matrix: [Technology Support](https://docs.speedscale.com/reference/technology-support/)
*   Shared proxymock proxy reference: [Language Configuration](https://docs.speedscale.com/proxymock/getting-started/language-reference/)

<a id="kubernetes-sidecar"></a>

## Kubernetes Sidecar

When Python runs with the Speedscale sidecar in `forward` or `dual` mode, configure the runtime to use the sidecar's forward proxy on `127.0.0.1:4140` unless you changed `proxy-out-port`.

Typical settings:

```bash
export HTTP_PROXY=http://127.0.0.1:4140
export HTTPS_PROXY=http://127.0.0.1:4140
```

If `tls-out` is enabled, also trust the Speedscale CA separately, commonly with `REQUESTS_CA_BUNDLE` for `requests`\-based applications.

See [Proxy Modes](https://docs.speedscale.com/getting-started/installation/sidecar/proxy-modes/) and [TLS Support](https://docs.speedscale.com/getting-started/installation/sidecar/tls/) for the shared sidecar behavior.

<a id="demo-app"></a>

## Demo App

*   Public demo: [speedscale/mock-lab](https://github.com/speedscale/mock-lab) (`languages/python` directory)
*   Stack: standard-library Python (no Flask, no Makefile) that calls one downstream, the CNCF projects API at `https://demo-api.trafficreplay.com`
*   Local run: `python3 app.py`
*   Traffic generator: `./lab/tests/run_tests.sh --recording`

This is the current public Python demo used for local proxymock examples.

<a id="proxymock"></a>

## proxymock

Use this path for the fastest Python first success on a developer workstation.

1.  1. Install and initialize proxymock
    
    ```
    brew install speedscale/tap/proxymock
    proxymock init
    ```
    
    Use browser sign-in by default. Use \`proxymock init --api-key <your key>\` only for CI or other headless environments.
    
2.  2. Clone the demo and start recording
    
    ```
    git clone https://github.com/speedscale/mock-lab
    cd mock-lab/languages/python
    proxymock record -- python3 app.py
    ```
    
    proxymock records the app while it starts the Python service as a child process. The app listens on port 8080 and calls the CNCF projects API downstream.
    
3.  3. Generate one real workflow
    
    ```
    ./lab/tests/run_tests.sh --recording
    ```
    
    Run the test driver from the repo root. It drives the requests that become the exported production-style trace.
    
4.  4. Stop the recording, then run with mocks
    
    ```
    cd mock-lab/languages/python
    proxymock mock -- python3 app.py
    ```
    
    The mocked run should no longer need live outbound dependencies.
    
5.  5. Replay the same traffic against a change
    
    ```
    cd mock-lab/languages/python
    proxymock replay --test-against http://localhost:8080
    ```
    
    Use replay as the regression check before shipping Python changes.
    

<a id="tls-trust"></a>

## TLS Trust

The demo app uses the standard-library `urllib` client and needs no manual CA configuration locally — proxymock injects the trusted bundle for you. For Python applications that use `requests`, trust the Speedscale certificate bundle with `REQUESTS_CA_BUNDLE`. See the shared [Language Configuration](https://docs.speedscale.com/proxymock/getting-started/language-reference/#tls-trust) page for the exact command and related options.