How to Organize Your AI Prompts

Use the following naming convention for all prompts in our library:

Format: [Category] - [Action] - [Specificity]

Examples:
- "Writing - Blog Post - Technical Tutorial"
- "Code - Review - Python Security Audit"
- "Analysis - Competitor - Quarterly Summary"
- "Support - Reply - Refund Request"

Rules:
1. Category must be one of: {{categories}}
2. Action should be a verb or verb phrase describing what the prompt does
3. Specificity narrows the use case so the prompt is findable
4. Maximum 60 characters total
5. Use title case, separated by hyphens

Apply this convention to the following prompt:
Title: "{{current_title}}"
Purpose: {{purpose}}

Suggested name:

Design a folder structure for organizing AI prompts based on my work context.

My roles/domains: {{domains}}
AI tools I use: {{tools}}
Types of prompts I write most: {{prompt_types}}
Team size: {{team_size}}

Create a folder structure that:
1. Has no more than 3 levels of nesting
2. Groups by use case first, not by AI model
3. Includes a "Templates" folder for reusable starting points
4. Includes an "Archive" folder for deprecated prompts
5. Uses clear, descriptive folder names (no abbreviations)

Return the structure as an indented tree with a brief description of what goes in each folder.

Folder structure:

Create a tagging taxonomy for a prompt library. Tags should be orthogonal to the folder structure (folders = what domain, tags = cross-cutting attributes).

Context: {{context}}

Design tags in these categories:

1. **Technique tags**: The prompting method used (e.g., few-shot, chain-of-thought, zero-shot, role-play, system-prompt)
2. **Status tags**: The prompt's lifecycle stage (e.g., draft, tested, production, deprecated)
3. **Model tags**: Which AI model it's optimized for (e.g., gpt-4, claude, gemini)
4. **Complexity tags**: How complex the prompt is (e.g., simple, intermediate, advanced)
5. **Output tags**: What the prompt produces (e.g., text, code, json, table, analysis)

For each category:
- List 4-6 recommended tags
- Explain when to use each one
- Note which tags should be required vs. optional

Taxonomy:

Create a version log entry for this prompt update.

Prompt name: {{prompt_name}}
Previous version: {{previous_version}}
New version: {{new_version}}

Generate a structured version log entry:

## Version {{version_number}} - {{date}}

### Changes
- [List each specific change made]

### Reason for update
[Why was this change needed? What problem did the previous version have?]

### Test results
- Model tested: [model name]
- Input used: [test input summary]
- Previous output quality: [1-5 rating + brief note]
- New output quality: [1-5 rating + brief note]

### Rollback notes
[Under what circumstances should we revert to the previous version?]

Write a brief guide for our team on how to share and reuse prompts effectively.

Team context: {{team_context}}
Tools we use: {{tools}}
Current pain points: {{pain_points}}

The guide should cover:

1. **When to share a prompt**: What qualifies a prompt for the shared library vs. keeping it personal?
2. **How to document a shared prompt**: What metadata must be included (description, example input/output, model requirements, known limitations)?
3. **How to request changes**: Process for suggesting improvements to shared prompts without breaking others' workflows
4. **Ownership and maintenance**: Who is responsible for keeping shared prompts updated?
5. **Naming and tagging standards**: Summary of our conventions (reference existing taxonomy)

Keep it under 500 words. Use bullet points. Make it practical, not theoretical.

Guide:

Audit my prompt library and identify what needs attention.

Here are my prompts:
{{prompt_list}}

For each prompt, evaluate:
1. **Last used**: Has it been used in the past 30 days? If not, flag for archive.
2. **Duplicates**: Are any prompts doing essentially the same thing? Suggest consolidation.
3. **Missing metadata**: Does it have a clear name, description, and tags? Flag incomplete entries.
4. **Outdated references**: Does it reference specific model versions or features that may have changed?
5. **Performance**: Based on the prompt structure, flag any that use known anti-patterns (vague instructions, no format specification, missing constraints).

Return a summary with:
- Total prompts reviewed
- Prompts to archive (with reason)
- Prompts to merge (which ones and why)
- Prompts needing metadata updates
- Prompts needing content updates
- Top 5 highest-priority actions

Audit results: