pyrestkit 1.2.2

Modern Python REST API Automation Toolkit With AI

🚀 PyRestKit

A modern Python framework for REST API automation with built-in validation, authentication, configuration management, and optional AI-assisted failure analysis.

---

Why PyRestKit?

Testing REST APIs often requires combining multiple libraries for:

• HTTP requests
• Authentication
• Configuration management
• Assertions
• JSON validation
• Logging
• Retry handling
• Environment management

As projects grow, these utilities become difficult to maintain consistently across teams.

PyRestKit provides a unified, extensible framework that brings these capabilities together while keeping your test code clean, maintainable, and type-safe.

In addition, PyRestKit offers optional AI-assisted failure analysis, enabling engineers to receive intelligent explanations and debugging suggestions from multiple Large Language Model (LLM) providers.

---

Features

REST API Client

• Simple request API
• GET / POST / PUT / PATCH / DELETE
• Custom headers
• Query parameters
• Cookies
• Multipart requests
• Request timeout support

---

Authentication

Built-in authentication helpers:

• Basic Authentication
• Bearer Token
• API Key
• OAuth 2.0 (yet to implement)
• Custom authentication strategies

---

Validation

Powerful response validation utilities.

Supported validations include:

• Status Code
• Headers
• JSON Schema
• Response Time
• JSON Path
• Custom Validators

---

Configuration

Configuration can be loaded from

• YAML
• Environment Variables
• Python Objects

Supports multiple environments including

• Development
• QA
• Staging
• Production

---

Assertions

Readable assertions for

• Status codes
• Headers
• JSON values
• Collections
• Response time
• Custom assertions

---

AI-Assisted Failure Analysis

PyRestKit can analyze failed API responses using AI.

Features include

• Root cause explanation
• Human-readable summaries
• Suggested fixes
• Optional system prompts
• Prompt templates
• Provider abstraction

AI support is completely optional and does not affect users who prefer traditional API testing.

---

Multiple AI Providers

PyRestKit currently supports:

Provider	Supported
OpenAI	✅
Anthropic	✅
Google Gemini	✅
Azure OpenAI	✅
Groq	✅
Cohere	✅
Mistral	✅
Ollama	✅
AWS Bedrock	✅

The provider abstraction allows additional providers to be implemented with minimal effort.

---

Installation

Standard Installation

pip install pyrestkit

---

Development Installation

git clone https://github.com/Raushanraj77/pyrestkit.git

cd pyrestkit

python -m venv .venv

source .venv/bin/activate

pip install -e .

---

Verify Installation

import pyrestkit

print(pyrestkit.__version__)

---

Quick Start

Create a Client

from pyrestkit import APIClient

client = APIClient(base_url="https://jsonplaceholder.typicode.com")

---

Send a Request

response = client.get("/posts/1")

---

Validate Response

response.validate_status_code(200)

response.validate_json()

response.validate_response_time(max_time_ms=1000)

---

Read JSON

data = response.json()

print(data["title"])

---

Authentication Example

Bearer Token

client = APIClient(base_url="https://example.com")

client.authenticate_bearer(token="your-token")

---

Basic Authentication

client.authenticate_basic(username="admin", password="secret")

---

API Key

client.authenticate_api_key(key="xxxxxxxx", header_name="X-API-Key")

---

Configuration Example

Load configuration from YAML.

base_url: https://api.example.com

timeout: 30

headers:
  Accept: application/json

from pyrestkit.config import Config

config = Config.from_yaml("config.yaml")

---

Response Validation

response.validate_status_code(200)

response.validate_header("Content-Type", "application/json")

response.validate_schema("schemas/user.json")

---

AI Failure Analysis

AI analysis can be enabled whenever deeper insight into a failed response is useful.

from pyrestkit.ai import AIAnalyzer
from pyrestkit.ai import AIConfig

config = AIConfig(provider="openai", model="gpt-4.1-mini", api_key="YOUR_API_KEY")

analyzer = AIAnalyzer(config)

analysis = analyzer.analyze(
    request=request,
    response=response,
)

print(analysis)

---

Supported AI Providers

PyRestKit uses a provider abstraction layer that allows switching AI providers without changing your test code.

                    AI Analyzer
                         │
                         ▼
                 Provider Factory
                         │
        ┌────────────────┼────────────────┐
        ▼                ▼                ▼
     OpenAI         Anthropic        Gemini
        │                │                │
        ▼                ▼                ▼
      Groq           Cohere         Azure OpenAI
        │
        ▼
     Mistral
        │
        ▼
     Ollama
        │
        ▼
    AWS Bedrock

Every provider implements the same interface.

class BaseAIProvider:
    def complete(
        self,
        prompt: str,
        *,
        config: AIConfig,
        system_prompt: str | None = None,
    ) -> str: ...

This makes it easy to:

• switch providers
• test providers
• implement custom providers
• support enterprise LLMs

---

Project Structure

pyrestkit/
│
├── pyrestkit/
│   │
│   ├── auth/
│   ├── client/
│   ├── config/
│   ├── exceptions/
│   ├── logging/
│   ├── retry/
│   ├── validators/
│   │
│   ├── ai/
│   │   ├── analyzer.py
│   │   ├── config.py
│   │   ├── exceptions.py
│   │   ├── factory.py
│   │   ├── prompt_loader.py
│   │   ├── prompts/
│   │   └── providers/
│   │
│   └── ...
│
├── docs/
├── examples/
├── tests/
├── pyproject.toml
└── README.md

---

Framework Architecture

                 Test

                  │

                  ▼

            API Client

                  │

                  ▼

        Authentication Layer

                  │

                  ▼

           HTTP Transport

                  │

                  ▼

            API Response

                  │

       ┌──────────┴──────────┐

       ▼                     ▼

 Validation             AI Analysis

       │                     │

       ▼                     ▼

 Assertions          Root Cause Analysis

---

Why AI Is Optional

PyRestKit is designed as an API automation framework first.

AI is an optional capability.

This provides several advantages:

• no AI dependency for normal users
• predictable execution
• no external API calls unless enabled
• works in offline environments
• enterprise friendly
• easier testing

If AI is never configured, PyRestKit behaves exactly like a traditional API automation framework.

---

Extending PyRestKit

Creating a custom AI provider requires implementing a single interface.

from pyrestkit.ai.providers.base import BaseAIProvider


class InternalProvider(BaseAIProvider):
    def complete(
        self,
        prompt: str,
        *,
        config,
        system_prompt=None,
    ) -> str:

        ...

        return result

Register the provider.

factory.register(
    "internal",
    InternalProvider,
)

Now it can be used exactly like built-in providers.

---

Example Workflow

response = client.post(
    "/users",
    json=payload,
)

response.validate_status_code(201)

response.validate_schema("schemas/user.json")

If validation fails:

analysis = analyzer.analyze(
    request=request,
    response=response,
)

print(analysis)

Example output:

Status Code Mismatch

Expected:
201

Received:
400

Likely Cause

The "email" field is required but missing
from the request payload.

Suggested Fix

Include a valid email address before
submitting the request.

---

Documentation

Detailed documentation is available in the docs directory.

Document	Description
getting-started.md	Installation and first project
architecture.md	Internal framework architecture
configuration.md	Configuration system
authentication.md	Authentication mechanisms
validation.md	Validators and assertions
ai.md	AI-assisted failure analysis
providers.md	AI provider implementations
examples.md	Complete usage examples
faq.md	Frequently asked questions
roadmap.md	Future development plans

---

Running Tests

Run all tests.

pytest

Run with coverage.

pytest --cov=pyrestkit

Run Ruff.

ruff check .

Run MyPy.

mypy .

Build package.

python -m build

---

Contributing

Contributions are welcome.

Please read:

• CONTRIBUTING.md
• CODE_OF_CONDUCT.md

before submitting pull requests.

---

Roadmap

Version 1.x

• REST API Client
• Authentication
• Validation
• Assertions
• Retry
• Configuration
• Logging
• AI Failure Analysis
• Multiple AI Providers

Version 2.x

Planned improvements include:

• Async HTTP Client
• HTML Reporting
• OpenAPI Integration
• AI Test Generation
• AI Assertion Suggestions
• Plugin System
• CLI
• VS Code Extension

---

Why PyRestKit?

PyRestKit aims to provide a clean, extensible, and modern approach to REST API automation.

Core principles:

• Simplicity
• Type Safety
• Extensibility
• Clean Architecture
• Optional AI
• Developer Experience

---

Requirements

• Python 3.10+
• requests
• pydantic
• jsonschema
• PyYAML

Optional:

• OpenAI
• Anthropic
• Gemini
• Azure OpenAI
• Groq
• Cohere
• Mistral
• Ollama
• AWS Bedrock

---

License

PyRestKit is licensed under the MIT License.

See the LICENSE file for details.

---

Support

If you encounter an issue:

• Open a GitHub Issue
• Start a GitHub Discussion
• Submit a Pull Request

---

Acknowledgements

PyRestKit is inspired by the design philosophies of several outstanding open-source projects, including:

• Requests
• HTTPX
• FastAPI
• Pydantic
• Playwright

while focusing specifically on delivering a modern, extensible framework for REST API automation.

---

Star the Project ⭐

If PyRestKit helps your team build reliable API automation, consider starring the repository to support the project and stay updated with future releases.