When you’re working on a team, managing a shared library of prompts can quickly become chaotic. Who wrote this prompt? Is it ready for production? Which prompts are for the billing system versus the support chatbot?

This guide provides a blueprint for using Prompt Lockbox’s organizational features to create a clean, collaborative, and easy-to-navigate prompt library for your team.

The Goal: To establish a clear, consistent, and scalable workflow for teams by leveraging:

Namespaces for organization.

  • The Status field for tracking a prompt’s lifecycle.
  • The Author field for ownership.

​
Organizing Prompts with Namespaces

The namespace field is a tool for bringing order to your library. Think of it like a folder path. By assigning a namespace to each prompt, you can group them by feature, domain, or team responsibility.

​
How to use ?

When creating a prompt, just define its namespace as a list of strings in your .yml file.

Example:

name: invoice-parser
namespace: [billing, invoices]
description: "Parses details from an uploaded invoice PDF."
...

Best Practice: Establish a consistent naming convention for your namespaces early on. For example, [feature, sub-feature] or [domain, action].

​
Visualizing the Structure with plb tree

Once your prompts are namespaced, the plb tree command becomes incredibly useful. It displays your library in a clean, hierarchical view, making it easy for anyone on the team to find what they need.

plb tree

​
Tracking the Prompt Lifecycle with status

The status field helps your team understand the maturity and intended use of each prompt. It prevents a half-finished “draft” prompt from accidentally being used in production.

​
The Standard Lifecycle

Here is a typical workflow using the status field:

1

1. Draft

When a prompt is first created, its status is Draft. This means it’s a work-in-progress and not ready for use.

status: "Draft"
2

2. In-Review

Once the author thinks the prompt is ready, they change the status to In-Review and open a Pull Request. This signals to other team members that it’s ready for peer review.

status: "In-Review"
3

3. Staging

After the review is approved, the prompt might be moved to Staging. This means it’s deployed to a testing or staging environment for final validation.

status: "Staging"
4

4. Production

Once it’s validated in staging, the prompt is moved to Production status and locked using plb lock. This is the final, trusted version.

status: "Production"
5

5. Deprecated/Archived

If a prompt is no longer needed or has been replaced by a newer version, its status can be set to Deprecated or Archived to signal that it should not be used in new development.

​
Viewing Statuses at a Glance

The plb list --wide command is perfect for getting a quick overview of the status of all prompts in the library.

plb list --wide

​
Set Author

In a team setting, knowing who to talk to about a specific prompt is essential. The author field solves this. When you run plb create or plb version in a Git repository, Prompt Lockbox automatically fills this field with your Git user name and email. Example:

author: "Ada lang <adalang@example.com"

By combining namespaces for organization, the status field for lifecycle tracking, and the author field for ownership, these features provide the structure needed to manage prompts effectively as your projects.