RunLog AI
ProductPricingDocsBlog
Documentation

Getting Started

  • Introduction
  • Quick Start
  • Installation

Core Concepts

  • Agent Runs
  • Policy Engine
  • Replay System

API Reference

  • Python SDK
  • TypeScript SDK
  • REST API

Python SDK Reference

Complete API reference for the RunLog AI Python SDK.

Installation
pip install runlog-ai

API Methods

RL()
Initialize a new RunLog client

Signature

RL(service: str, env: str, api_key: str = None)

Example

rl = RL(
    service="my-agent",
    env="production",
    api_key="rl_..."
)
rl.run()
Create a new agent run context

Signature

run(task: str, metadata: dict = None)

Example

with rl.run(task="user-query", metadata={"user_id": "123"}):
    # Your agent logic here
    pass
rl.tool()
Track external tool or API calls

Signature

tool(name: str, **kwargs)

Example

results = rl.tool(
    "search_api",
    query="python tutorial",
    max_results=10
)
rl.enforce()
Apply runtime policy enforcement

Signature

enforce(policy: str, **kwargs)

Example

rl.enforce(
    "max_cost_per_run",
    limit=1.00,
    currency="USD"
)
Configuration Options
Available configuration parameters

Environment Variables

RUNLOG_API_KEYYour API key
RUNLOG_BASE_URLAPI base URL

Client Options

timeoutRequest timeout
retry_countRetry attempts
RunLog AI

RunLog AI records every agent step, enforces policies before risky actions, and replays runs to fix issues—before users are impacted.

Product

  • Features
  • Pricing
  • Documentation
  • API Reference

Company

  • About
  • Blog
  • Careers
  • Contact

Legal

  • Privacy Policy
  • Terms of Service
  • Security
  • Status

© 2025 RunLog AI. All rights reserved.

Built with Next.js and Tailwind CSS