Agent Academy

Agent Academy is a free, source-linked library of skills, prompts, MCP servers and bot templates, organised into courses by profession, for people who hand real work to AI agents: 1,467 live resources, each with its original link, license and safety tags.

Skill

temporal-python-testing

Test Temporal workflows with pytest, time-skipping, and mocking strategies. Covers unit testing, integration testing, replay testing, and local development setup. Use when implementing Temporal workflow tests or debugging test failures.

No credentialsDoes not send

https://github.com/wshobson/agents/blob/main/plugins/backend-development/skills/temporal-python-testing/SKILL.md

Install
npx skills add wshobson/agents --skill temporal-python-testing

Original content

Full originalMITSource

Republished in full under its license, with attribution. The source is authoritative.

name: temporal-python-testing
description: Test Temporal workflows with pytest, time-skipping, and mocking strategies. Covers unit testing, integration testing, replay testing, and local development setup. Use when implementing Temporal workflow tests or debugging test failures.

Temporal Python Testing Strategies

Comprehensive testing approaches for Temporal workflows using pytest, progressive disclosure resources for specific testing scenarios.

When to Use This Skill

  • Unit testing workflows - Fast tests with time-skipping
  • Integration testing - Workflows with mocked activities
  • Replay testing - Validate determinism against production histories
  • Local development - Set up Temporal server and pytest
  • CI/CD integration - Automated testing pipelines
  • Coverage strategies - Achieve ≥80% test coverage

Testing Philosophy

Recommended Approach (Source: docs.temporal.io/develop/python/testing-suite):

  • Write majority as integration tests
  • Use pytest with async fixtures
  • Time-skipping enables fast feedback (month-long workflows → seconds)
  • Mock activities to isolate workflow logic
  • Validate determinism with replay testing

Three Test Types:

  1. Unit: Workflows with time-skipping, activities with ActivityEnvironment
  2. Integration: Workers with mocked activities
  3. End-to-end: Full Temporal server with real activities (use sparingly)

Available Resources

This skill provides detailed guidance through progressive disclosure. Load specific resources based on your testing needs:

Unit Testing Resources

File: resources/unit-testing.md

When to load: Testing individual workflows or activities in isolation

Contains:

  • WorkflowEnvironment with time-skipping
  • ActivityEnvironment for activity testing
  • Fast execution of long-running workflows
  • Manual time advancement patterns
  • pytest fixtures and patterns

Integration Testing Resources

File: resources/integration-testing.md

When to load: Testing workflows with mocked external dependencies

Contains:

  • Activity mocking strategies
  • Error injection patterns
  • Multi-activity workflow testing
  • Signal and query testing
  • Coverage strategies

Replay Testing Resources

File: resources/replay-testing.md

When to load: Validating determinism or deploying workflow changes

Contains:

  • Determinism validation
  • Production history replay
  • CI/CD integration patterns
  • Version compatibility testing

Local Development Resources

File: resources/local-setup.md

When to load: Setting up development environment

Contains:

  • Docker Compose configuration
  • pytest setup and configuration
  • Coverage tool integration
  • Development workflow

Quick Start Guide

Basic Workflow Test

import pytest
from temporalio.testing import WorkflowEnvironment
from temporalio.worker import Worker

@pytest.fixture
async def workflow_env():
    env = await WorkflowEnvironment.start_time_skipping()
    yield env
    await env.shutdown()

@pytest.mark.asyncio
async def test_workflow(workflow_env):
    async with Worker(
        workflow_env.client,
        task_queue="test-queue",
        workflows=[YourWorkflow],
        activities=[your_activity],
    ):
        result = await workflow_env.client.execute_workflow(
            YourWorkflow.run,
            args,
            id="test-wf-id",
            task_queue="test-queue",
        )
        assert result == expected

Basic Activity Test

from temporalio.testing import ActivityEnvironment

async def test_activity():
    env = ActivityEnvironment()
    result = await env.run(your_activity, "test-input")
    assert result == expected_output

Coverage Targets

Recommended Coverage (Source: docs.temporal.io best practices):

  • Workflows: ≥80% logic coverage
  • Activities: ≥80% logic coverage
  • Integration: Critical paths with mocked activities
  • Replay: All workflow versions before deployment

Key Testing Principles

  1. Time-Skipping - Month-long workflows test in seconds
  2. Mock Activities - Isolate workflow logic from external dependencies
  3. Replay Testing - Validate determinism before deployment
  4. High Coverage - ≥80% target for production workflows
  5. Fast Feedback - Unit tests run in milliseconds

How to Use Resources

Load specific resource when needed:

  • "Show me unit testing patterns" → Load resources/unit-testing.md
  • "How do I mock activities?" → Load resources/integration-testing.md
  • "Setup local Temporal server" → Load resources/local-setup.md
  • "Validate determinism" → Load resources/replay-testing.md

Additional References

  • Python SDK Testing: docs.temporal.io/develop/python/testing-suite
  • Testing Patterns: github.com/temporalio/temporal/blob/main/docs/development/testing.md
  • Python Samples: github.com/temporalio/samples-python

Related

SkillSoftware engineering

claude-api

Reference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`,…

Sourcegithub.com/anthropics/skills/blob/main/skills/claude-api/SKILL.md

No credentialsDoes not sendApache-2.0
SkillSoftware engineering

doc-coauthoring

Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar…

Sourcegithub.com/anthropics/skills/blob/main/skills/doc-coauthoring/SKILL.md

No credentialsDoes not sendnot stated
SkillSoftware engineering

mcp-builder

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

Sourcegithub.com/anthropics/skills/blob/main/skills/mcp-builder/SKILL.md

No credentialsDoes not sendApache-2.0
SkillSoftware engineering

skill-creator

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

Sourcegithub.com/anthropics/skills/blob/main/skills/skill-creator/SKILL.md

No credentialsDoes not sendApache-2.0
SkillSoftware engineering

webapp-testing

Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.

Sourcegithub.com/anthropics/skills/blob/main/skills/webapp-testing/SKILL.md

No credentialsDoes not sendApache-2.0
SkillSoftware engineering

ai-debt-detector

Use after generating code, after accepting AI suggestions, or when reviewing AI-written modules. Also use when code works but feels brittle, when error handling seems thin, when orphaned resources or missing cleanup are suspected, or when the agent claims done but hidden debt may exist. Catches the specific failure patterns AI agents produce that humans would not.

Sourcegithub.com/wshobson/agents/blob/main/plugins/skill-forge-essentials/skills/ai-debt-detector/SKILL.md

No credentialsDoes not sendMIT