ToggleMesh Python SDK (Local Evaluation)

A lightweight, robust Python SDK for evaluating feature flags with ToggleMesh.

This SDK uses Local Evaluation. It connects to the ToggleMesh control plane once to fetch all rules, establishes an SSE (Server-Sent Events) connection to listen for updates, and evaluates all flags locally in-memory without making network requests. This ensures ultra-fast responses (sub-millisecond) suitable for high-throughput backend services.

Installation

pip install togglemesh

Usage

import time
from togglemesh import ToggleMeshClient, ToggleMeshOptions

client = ToggleMeshClient(ToggleMeshOptions(
    base_url="http://localhost:5000",
    server_key="YOUR_API_KEY"
))

# 1. Check a flag locally (Zero network latency!)
# Pass identity and context per-evaluation, since the server handles multiple users concurrently.
if client.is_enabled("new-feature", default_value=False, identity="user-123", tenant="sample_corp"):
    print("Feature is ON")
else:
    print("Feature is OFF")

# 2. Stop background threads when the app shuts down
client.stop()

Supported Features

• Percentage Rollouts (Murmur3 / FNV-1a hashing)
• Operator matching (Strings, Numbers, Dates, SemVer, Regex)
• Real-time SSE updates
• Contextual Rollouts
• Segment Evaluation
• JSON Parsing client.get_json(...)
• Analytics Event Tracking (A/B testing conversions):

  client.track("checkout_completed", value=150.0, identity="user-123", cart_size=3)
  

🚀 CLI Tool (Code Generation)

The togglemesh Python package also includes a powerful Command Line Interface for synchronizing feature flags from the control plane and generating strongly-typed Python classes. This eliminates typos and gives you full IDE autocomplete for your flag keys!

1. Setup Configuration

Run the configuration wizard to link the CLI to your ToggleMesh environment.

togglemesh config

2. Generate Typed Flags

Synchronize your flags and output a native Python file containing your flag constants:

togglemesh sync -l python -o ./flags.py

Now you can use the generated Flags class in your code:

from flags import Flags
from togglemesh import ToggleMeshClient, ToggleMeshOptions

# ...
client.is_enabled(Flags.NEW_CHECKOUT_FLOW, default_value=False, identity="user-123")