Skip to content

Commit 9e3ef03

Browse files
committed
update content
1 parent 03ae020 commit 9e3ef03

21 files changed

Lines changed: 526 additions & 572 deletions
Lines changed: 5 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -1,15 +1,15 @@
11
### YamlMime:ModuleUnit
22
uid: learn.wwl.prompt-versioning-genaiops.version-control-for-prompts
3-
title: Apply version control principles to prompts
3+
title: Apply version control to prompts
44
metadata:
5-
title: Apply version control principles to prompts
6-
description: "Learn to apply version control principles to manage prompts as code assets"
7-
ms.date: 01/07/2026
5+
title: Apply version control to prompts
6+
description: "Learn why prompts need version control and how DevOps practices apply to prompt management"
7+
ms.date: 01/26/2026
88
author: madiepev
99
ms.author: madiepev
1010
ms.topic: unit
1111
ms.custom:
1212
- N/A
13-
durationInMinutes: 8
13+
durationInMinutes: 5
1414
content: |
1515
[!include[](includes/2-version-control-for-prompts.md)]
Lines changed: 15 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,15 @@
1+
### YamlMime:ModuleUnit
2+
uid: learn.wwl.prompt-versioning-genaiops.foundry-agents-prompt-versioning
3+
title: Understand Microsoft Foundry agents and prompt versioning
4+
metadata:
5+
title: Understand Microsoft Foundry agents and prompt versioning
6+
description: "Learn how prompts integrate with Microsoft Foundry agent definitions and how to version agents using Git tags"
7+
ms.date: 01/26/2026
8+
author: madiepev
9+
ms.author: madiepev
10+
ms.topic: unit
11+
ms.custom:
12+
- N/A
13+
durationInMinutes: 7
14+
content: |
15+
[!include[](includes/3-foundry-agents-prompt-versioning.md)]

learn-pr/wwl-data-ai/prompt-versioning-genaiops/3-github-repository-structure.yml

Lines changed: 0 additions & 15 deletions
This file was deleted.
Lines changed: 15 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,15 @@
1+
### YamlMime:ModuleUnit
2+
uid: learn.wwl.prompt-versioning-genaiops.github-repository-structure
3+
title: Organize prompts in GitHub repositories
4+
metadata:
5+
title: Organize prompts in GitHub repositories
6+
description: "Learn how to organize prompts in GitHub repositories with proper folder structure, file formats, and naming conventions"
7+
ms.date: 01/26/2026
8+
author: madiepev
9+
ms.author: madiepev
10+
ms.topic: unit
11+
ms.custom:
12+
- N/A
13+
durationInMinutes: 10
14+
content: |
15+
[!include[](includes/4-github-repository-structure.md)]

learn-pr/wwl-data-ai/prompt-versioning-genaiops/4-prompt-workflow-development.yml

Lines changed: 0 additions & 15 deletions
This file was deleted.
Lines changed: 15 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,15 @@
1+
### YamlMime:ModuleUnit
2+
uid: learn.wwl.prompt-versioning-genaiops.prompt-workflow-development
3+
title: Develop safe prompt deployment workflows
4+
metadata:
5+
title: Develop safe prompt deployment workflows
6+
description: "Learn to develop workflows for testing and deploying prompts safely through development, validation, and production stages"
7+
ms.date: 01/26/2026
8+
author: madiepev
9+
ms.author: madiepev
10+
ms.topic: unit
11+
ms.custom:
12+
- N/A
13+
durationInMinutes: 8
14+
content: |
15+
[!include[](includes/5-prompt-workflow-development.md)]

learn-pr/wwl-data-ai/prompt-versioning-genaiops/5-exercise.yml renamed to learn-pr/wwl-data-ai/prompt-versioning-genaiops/6-exercise.yml

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -12,4 +12,4 @@ metadata:
1212
- N/A
1313
durationInMinutes: 15
1414
content: |
15-
[!include[](includes/5-exercise.md)]
15+
[!include[](includes/6-exercise.md)]
Lines changed: 25 additions & 25 deletions
Original file line numberDiff line numberDiff line change
@@ -1,10 +1,10 @@
11
### YamlMime:ModuleUnit
22
uid: learn.wwl.prompt-versioning-genaiops.knowledge-check
3-
title: Module assessment
3+
title: Knowledge check
44
metadata:
5-
title: Module assessment
5+
title: Knowledge check
66
description: "Knowledge check for prompt versioning module"
7-
ms.date: 01/07/2026
7+
ms.date: 01/26/2026
88
author: madiepev
99
ms.author: madiepev
1010
ms.topic: unit
@@ -15,36 +15,36 @@ durationInMinutes: 3
1515
quiz:
1616
title: ""
1717
questions:
18-
- content: "Why is version control important for AI prompts in production environments?"
18+
- content: "What happens when you create or update an agent in Microsoft Foundry?"
1919
choices:
20-
- content: "It allows you to track changes, roll back problematic prompts, and maintain consistency across deployments."
21-
isCorrect: true
22-
explanation: "Correct. Version control provides essential capabilities for managing prompt changes safely in production."
23-
- content: "It automatically improves prompt performance through machine learning."
20+
- content: "The existing agent version is overwritten with no history."
2421
isCorrect: false
25-
explanation: "Incorrect. Version control doesn't automatically improve prompt performance - it provides management capabilities."
26-
- content: "It eliminates the need for testing prompts before deployment."
22+
explanation: "Incorrect. Microsoft Foundry preserves agent version history."
23+
- content: "A new agent version is created automatically, whether you use the portal or the Python SDK."
24+
isCorrect: true
25+
explanation: "Correct. Microsoft Foundry creates a new version whenever you create or update an agent, maintaining version history."
26+
- content: "Version numbers must be manually assigned in your code."
2727
isCorrect: false
28-
explanation: "Incorrect. Version control complements testing but doesn't replace the need for thorough prompt validation."
29-
- content: "Which GitHub repository structure best supports prompt versioning for GenAIOps workflows?"
28+
explanation: "Incorrect. Microsoft Foundry automatically assigns incremental version numbers."
29+
- content: "In a proper prompt workflow, what should happen before a prompt change reaches production?"
3030
choices:
31-
- content: "A single file containing all prompts without any organization"
31+
- content: "Changes should be deployed immediately to get real user feedback."
3232
isCorrect: false
33-
explanation: "Incorrect. A single unorganized file makes it difficult to manage multiple prompts and their versions."
34-
- content: "Organized folders with clear naming conventions, separate files for each prompt, and documentation"
33+
explanation: "Incorrect. Deploying directly to production skips critical validation steps."
34+
- content: "Changes should be tested in development, validated in staging, reviewed by the team, and then deployed with monitoring."
3535
isCorrect: true
36-
explanation: "Correct. A well-structured repository with clear organization supports better prompt management and collaboration."
37-
- content: "Binary files that cannot be easily read or compared"
36+
explanation: "Correct. This systematic approach ensures changes are validated at each stage before impacting users."
37+
- content: "Only the person who wrote the prompt needs to approve it."
3838
isCorrect: false
39-
explanation: "Incorrect. Binary files don't support the transparency and diff capabilities needed for effective version control."
40-
- content: "What is the recommended approach for testing prompt changes before production deployment?"
39+
explanation: "Incorrect. Team review provides valuable perspective and catches issues the author might miss."
40+
- content: "What do Git tags provide for agent deployments?"
4141
choices:
42-
- content: "Test directly in production to get real user feedback"
42+
- content: "They automatically deploy agents to production."
4343
isCorrect: false
44-
explanation: "Incorrect. Testing directly in production introduces unnecessary risks and potential user impact."
45-
- content: "Create a branch, test the prompt changes, document performance differences, and use pull requests for review"
44+
explanation: "Incorrect. Git tags mark versions but don't trigger deployments by themselves."
45+
- content: "They create a traceable relationship between repository versions and deployed agents, simplifying debugging and rollback."
4646
isCorrect: true
47-
explanation: "Correct. This approach follows software engineering best practices and enables safe prompt iteration."
48-
- content: "Make changes without testing since prompts are just text"
47+
explanation: "Correct. Git tags enable you to quickly identify which repository state corresponds to any deployed agent version."
48+
- content: "They replace the need for commit messages."
4949
isCorrect: false
50-
explanation: "Incorrect. Prompt changes can significantly impact AI behavior and should always be tested thoroughly."
50+
explanation: "Incorrect. Tags complement commit messages by marking specific deployment milestones."
Lines changed: 1 addition & 15 deletions
Original file line numberDiff line numberDiff line change
@@ -1,20 +1,6 @@
11

22
Quick prompt changes that break AI agents in production are hard to debug. Many teams struggle with managing prompt versions safely, leading to unexpected behavior, and difficult rollbacks.
33

4-
## Common prompt management challenges
5-
6-
Teams working with AI agents frequently encounter these critical issues:
7-
8-
**Version tracking problems**: Multiple team members edit prompts without coordination, creating confusion about which version is currently deployed. Changes accumulate over time without documentation, making it impossible to identify what caused performance degradation.
9-
10-
**Collaboration difficulties**: Different departments (marketing, engineering, customer service) need to update prompts, but there's no approval workflow. Critical changes go directly to production without peer review or testing.
11-
12-
**Rollback challenges**: When prompts break in production, teams can't quickly revert to the last working version. Emergency fixes become manual efforts to recreate stable prompts from memory or scattered documentation.
13-
14-
**Testing gaps**: Prompt changes deploy without validation against real-world scenarios. Teams discover issues only after customers experience problems, leading to decreased satisfaction and trust.
15-
16-
**Knowledge management**: When team members leave, institutional knowledge about prompt design decisions and effective variations disappears. New team members struggle to understand why specific prompts work or how to modify them safely.
17-
184
## Scenario
195

206
Consider a customer service team at a software company that uses Microsoft Foundry to power their AI chat agent. Last month, someone updated the agent's system prompt to sound more friendly, but it caused the agent to give inconsistent responses to technical questions. The team couldn't figure out exactly what changed or quickly revert to the previous working version. Customer satisfaction dropped, and it took days to manually recreate a stable prompt from memory.
@@ -23,4 +9,4 @@ This scenario illustrates multiple challenges: lack of version control, missing
239

2410
## What you learn
2511

26-
In this module, you learn how to apply version control principles to AI agent prompts. You learn one approach to design a GitHub repository structure for collaboration. And finally you explore how to develop a safe workflow for testing and deploying prompt changes in Microsoft Foundry.
12+
In this module, you learn how to work with GitHub and Microsoft Foundry to version and deploy AI agent prompts. You explore one approach for organizing your repository structure for collaboration. While multiple organizational strategies exist, this example demonstrates core concepts you can adapt to your needs. You also learn how to develop a safe workflow for testing and deploying prompt changes.
Lines changed: 20 additions & 31 deletions
Original file line numberDiff line numberDiff line change
@@ -1,62 +1,51 @@
1-
It's 2 AM. Your AI customer service agent just went haywire. Customers are getting rude responses instead of helpful ones. The team scrambles to figure out what changed. Was it the model? A bug in the code? No. Someone made a *tiny prompt tweak* three hours ago to sound more casual. Now you're manually reverting changes from memory while angry customers flood social media.
1+
Prompt changes in production AI systems require the same systematic management approach as traditional software code changes.
22

3-
This nightmare scenario happens when teams treat prompts like throwaway text instead of what they really are: critical system components that determine AI behavior.
3+
In the customer service team scenario, someone updated the agent's system prompt without any tracking or approval process. The team couldn't identify what changed, who made the change, or how to quickly restore the working version. This chaos could be prevented with proper version control.
4+
5+
Here, you learn why prompts are production-critical assets, how established DevOps practices apply to prompt management, and what problems arise without systematic version control.
46

57
## Understand why prompts need the same care as code
68

79
Every prompt you write in Microsoft Foundry becomes a live configuration controlling how your AI system behaves. When users interact with your agent, your prompt shapes every response: affecting accuracy, tone, safety, and even cost.
810

9-
Unlike traditional software where code goes through compilation and testing, prompt changes take effect instantly. Change one word, and you immediately change how thousands of users experience your system. This makes prompts production-critical assets that deserve the same careful management as any other system component.
11+
Unlike traditional software where code goes through compilation and testing, prompt changes take effect instantly. Change one word, and you immediately change how thousands of users experience your system. These factors make prompts production-critical assets that deserve the same careful management as any other system component.
1012

11-
## Apply Development Operations practices for prompt management
13+
## Apply Development Operations (DevOps) practices for prompt management
1214

1315
Software teams solved similar challenges decades ago. Here's how established Development Operations (DevOps) practices translate to prompt management:
1416

15-
### Source control and versioning
16-
17-
**What it is**: Track every change to files with complete history, so you can see what changed, when, and why.
18-
19-
**For prompts**: Version every prompt iteration so you can compare changes, understand evolution, and revert when needed. Instead of editing prompts directly in Foundry, you store them in GitHub with full change history.
20-
21-
### Code reviews and team validation
22-
23-
**What it is**: Require team approval before changes reach production to catch issues early.
24-
25-
**For prompts**: Review prompt changes before deployment to verify they maintain accuracy, safety, and brand voice. Multiple eyes on changes prevent "small tweaks" from breaking system behavior.
26-
27-
### Environment separation (dev/staging/prod)
28-
29-
**What it is**: Test changes in isolated environments before they affect real users.
30-
31-
**For prompts**: Validate prompt changes in development environments before deploying to production. This prevents untested prompts from reaching customers.
32-
33-
### Rollback capabilities
34-
35-
**What it is**: Quickly revert to previous working versions when problems arise.
36-
37-
**For prompts**: Restore previous prompt versions instantly when issues occur. No more reconstructing prompts from memory during outages.
17+
| DevOps Practices | Traditional Software | For AI Prompts |
18+
|-----------------|---------------------|----------------|
19+
| **Source control and versioning** | Track every change to files with complete history, so you can see what changed, when, and why | Version every prompt iteration so you can compare changes, understand evolution, and revert when needed. Instead of editing prompts directly in Foundry, you store them in GitHub with full change history |
20+
| **Code reviews and team validation** | Require team approval before changes reach production to catch issues early | Review prompt changes before deployment to verify they maintain accuracy, safety, and brand voice. Multiple eyes on changes prevent "small tweaks" from breaking system behavior |
21+
| **Environment separation** | Test changes in isolated environments before they affect real users | Validate prompt changes in development environments before deploying to production. Environment separation prevents untested prompts from reaching customers |
22+
| **Roll back capabilities** | Quickly revert to previous working versions when problems arise | Restore previous prompt versions instantly when issues occur. No more reconstructing prompts from memory during outages |
3823

3924
## Identify what goes wrong without proper management
4025

4126
Without proper management, predictable failures emerge:
4227

28+
> [!WARNING]
29+
> These common failures can severely harm your AI system's reliability and user trust.
30+
4331
- **Silent degradation**: A "harmless" wording change reduces accuracy across scenarios, but you don't notice until users complain
4432
- **Environment drift**: Development works perfectly, but production behaves differently because someone updated the prompt in one environment but not the other
4533
- **Crisis recovery**: Production breaks, but you can't restore the working version because changes weren't tracked
46-
- **Lost knowledge**: Months later, you remove "unnecessary" instructions that were actually critical safety measures
34+
- **Lost knowledge**: Months later, you remove "unnecessary" instructions that were critical safety measures
4735

4836
## Evaluate how proper prompt management solves these problems
4937

5038
Version control transforms prompt chaos into systematic operations:
5139

5240
- **Intentional changes**: Every modification goes through review and validation
5341
- **Traceable problems**: When issues arise, you can see exactly what changed and when
54-
- **Instant recovery**: Rollback to working versions in seconds, not hours
42+
- **Instant recovery**: Roll back to working versions in seconds, not hours
5543
- **Team coordination**: Everyone works from the same prompt versions across all environments
5644
- **Audit trail**: Complete history of who changed what and why
5745

58-
This systematic approach gives you confidence to iterate quickly while maintaining reliability. Changes become improvements, not risks.
46+
> [!TIP]
47+
> This systematic approach gives you confidence to iterate quickly while maintaining reliability. Changes become improvements, not risks.
5948
6049
## Implement GitHub for prompt management
6150

62-
You understand why prompts need version control. In the next unit, you'll learn how to use GitHub to implement proper prompt management. From storing prompts in repositories to managing changes through pull requests. You explore how familiar GitHub workflows adapt perfectly to prompt versioning needs.
51+
You understand why prompts need version control. In the next unit, you'll learn how to use GitHub to implement proper prompt management. From storing prompts in repositories, to managing changes through pull requests. You explore how familiar GitHub workflows adapt perfectly to prompt versioning needs.

0 commit comments

Comments
 (0)