Help-center article from a cluster of resolved tickets
Turns several resolved tickets about the same problem into a self-service help article with a symptom-based title, steps, troubleshooting and a clear contact path. For support leads reducing repeat tickets.
Prompt in Customer support by AISkillHub Library
What it does
When the same question shows up in the queue every week, the fix is usually a help article. Writing it from memory tends to produce an article organised around how the product works internally rather than what the customer sees. This prompt builds the article from real resolved tickets, so it uses the customer's words for the problem and covers the variations that actually occur.
**Use it when** you have three or more resolved tickets about one issue and want a draft article for your help center or knowledge base.
**What you get** - A title written as the customer would search ("Can't log in after changing your email"), plus two alternative titles. - A one-paragraph summary of who this affects and why it happens, in plain words. - Numbered resolution steps, with the most common fix first. - A troubleshooting section for the variations found in the tickets. - A "contact us if" section that says exactly what information to include, so escalations arrive complete. - A list of search keywords and phrasings taken from the tickets. - A short note to the support lead on gaps: cases in the tickets the article does not resolve, and anything that looks like a product bug worth reporting.
**What it will not do** - It will not describe UI labels, menu paths or settings that do not appear in the tickets or the product notes you provide. Where it needs a path it cannot confirm, it writes `[confirm menu path]`. - It will not include customer names, emails or account details from the tickets. - It will not promise fixes or release dates.
Remove personal data from the tickets before pasting if your policies require it.
How to use it
## How to use
1. Open `prompts/help-article-from-tickets.md` and copy everything below the title line. 2. Replace each `{{variable}}` with your own content. Paste long material (notes, transcripts, tables) directly in place of the variable; keep the surrounding tags so the model can tell inputs apart. 3. Start a new chat in ChatGPT, Claude or Gemini and paste the filled prompt as your first message. 4. Read the output against the quality checks at the end of the prompt before you use or send it.
## Variables
- `{{tickets}}`: three or more resolved tickets about the same issue, including the agent's resolution. Separate each with `---`. - `{{product_notes}}`: accurate UI labels, menu paths or settings for this area, if you have them. Optional. - `{{product_name}}`: your product name. - `{{reading_level}}`: e.g. `general consumers`, `IT administrators`.
If you do not have a value for an optional variable, write `none` rather than deleting it, so the prompt still reads correctly.
## Reusing it
To reuse it often, save the prompt (with the variables left in) as the instructions of a Claude Project, a custom GPT or a Gemini Gem, then paste only the inputs in each new chat.
Files
Save each file at the path shown, relative to your project root.