The Prompt Library is RankGear’s collection of reusable AI prompts — both the built-in defaults and any you write yourself — that you can edit, save, filter, pick from, and delete.
| Term | Prompt Library |
|---|---|
| Category | AI, Content Generation and Experiments |
| Also known as | Saved Prompts |
| Where it appears | Supplemental Content |
What it means in RankGear
The Prompt Library lives in the Supplemental Content area, where RankGear’s AI features draw on it. It holds two kinds of entry: the default prompts that ship with the app for common jobs like critiquing an experiment or drafting a section, and the custom prompts you create and keep for your own recurring tasks. Every entry is editable, so a shipped default can be tuned to your voice and saved as your own version. The library is the shared shelf those prompts sit on — you filter it to narrow the list, select an entry to run it, and delete anything you no longer need.
How to interpret it
Treat the library as the input side of any AI action, not the output. Before you run a prompt, read the context it will send, the model it is set to use, and the cost that model choice implies, then look at what the prompt is actually asking the AI to produce. A saved prompt is only as good as the instructions inside it, so a stale or over-broad prompt will quietly shape every answer it generates. Editing and re-saving the prompt is usually a better fix than correcting each answer after the fact.
Example
You keep a custom prompt called “Experiment critique” that feeds a running test’s setup into the AI and asks it to flag confounds and weak isolation. You open the Prompt Library, filter to your saved prompts, select that one, and run it against your current experiment. The AI returns a critique; you read it, correct the two claims it got wrong, and keep the rest. A week later you tighten the prompt’s wording and save it, so the next run starts from the improved version.
Important considerations
- AI output can be incomplete or wrong. Verify factual claims, credentials, statistics, citations, and experimental isolation before you act on a generated answer.
- Editing a default prompt changes what it produces everywhere you run it — save a copy if you want to keep the original behavior alongside your version.
- Model choice carries a cost and a quality trade-off; check which model a prompt is set to before running it at scale.
- The library organizes and stores prompts — it does not judge their quality. A saved prompt is a convenience, not a guarantee of a good result.
Related terms
Part of the RankGear glossary · how RankGear measures · the 870 factors.