Prompt Library PDF: A Printable Collection You Can Adapt

A printable prompt library records your most useful AI prompts in a format any team member can pick up, fill in, and use without guessing what inputs are needed. The PDF or printed version is a snapshot for sharing; a living tool like fenbs keeps the current approved version updated after the snapshot is out of date.

The templates below are organized as three sections you can copy into a document: writing prompts, research and analysis prompts, and project management prompts. Each entry uses a consistent format with fields for the prompt text, required inputs, and an output check. Copy the whole set or take the sections that fit your team's work.

Format for each prompt entry

Every entry in a printable library should answer four questions: what does this prompt do, what inputs does it need, what should the output include, and what should you check before using the result. A consistent format across all entries makes the library faster to scan and easier to hand off to someone new. Decide on the format before you start adding entries, because reformatting a large library later is tedious.

• Name: a short label for the task, such as "First draft" or "Gap finder".
• Prompt text: the full instruction with [bracketed placeholders] for inputs that change between uses.
• Required inputs: the information the user must supply before running the prompt.
• Output check: one or two things to verify before using the output, such as "confirm no claims were added that are not in the source material."
• Notes: when this prompt works well, when it does not, and any alternatives to try if results are inconsistent.

Writing prompts

Task: Write a [document type] about [topic].
Source: [paste source material or write "use the notes above"].
Audience: [reader description].
Tone: [formal / conversational / technical].
Length: [word count or page limit].
End with a one-sentence summary of the main point.

Required inputs: document type, topic, source, audience, tone, length.
Output check: confirm the summary matches the document and no facts were added beyond the source material.

Task: Review the draft above.
Identify [number] things a [target reader] might find unclear or missing.
Do not rewrite the draft.
List each issue as one sentence, then suggest what would fix it.

Required inputs: the draft (paste above), number of issues, target reader.
Output check: confirm suggestions address clarity or completeness, not just style preference.

When you adapt these prompts for your team, keep the "required inputs" line even if it seems obvious. Someone picking up the library six months from now will appreciate knowing exactly what to fill in before running the prompt, rather than having to infer it from the placeholders.

Research and analysis prompts

Task: Read the document above.
List the [number] questions most likely to come from a reviewer who has not seen the source material.
For each, say whether the document answers it, answers it partially, or leaves it open.

Required inputs: the document (paste above), number of questions.
Output check: confirm questions are genuine gaps, not questions the document already answers clearly.

Task: Read the claims in the section above.
Identify any that need an external source to be credible.
For each claim, write it as a short sentence and note what kind of source would satisfy it.
Do not invent sources.

Required inputs: the section (paste above).
Output check: confirm the prompt did not add new claims that are not in the original text.

The source checker is particularly useful before publishing anything that makes factual claims about other products, regulations, or statistics. Running it before review catches the claims that need a citation, which is faster than finding them after the document has been shared.

These two prompts work well in sequence. Run the gap finder to find what is missing, then the source checker to find what is unsubstantiated. Address the gaps first, since a gap in the argument is harder to fix than a missing citation.

Project management prompts

Task: Write a status update for [project name] using only the task notes above.
Audience: [internal team / client / stakeholder].
Sections:
- Completed: 2-3 bullet points
- In progress: 2-3 bullet points with current status
- Blocked or waiting: what is stuck and who needs to act
- Next: the single most important thing to do next
Keep each bullet to one sentence. Do not add context that is not in the notes.

Required inputs: task notes (paste above), project name, audience.
Output check: confirm no internal notes or cost figures appeared in a client-facing update.

Task: Write a decision record for the choice described in the notes above.
Include:
- Decision: one sentence stating what was decided
- Reason: one to three sentences explaining why
- Who decided: [name or role]
- Date: [date]
- What this replaces: any earlier decision this supersedes
- What this does not cover: related questions that are still open

Required inputs: the notes (paste above), who decided, date.
Output check: confirm the decision is stated as a fact, not as a recommendation.

When project context and decisions live in fenbs, ask your AI to read that context before running these prompts. The status update will be more accurate if the AI can see the current task states directly rather than relying on notes you pasted in.

The limits of a static PDF

A printed or PDF prompt library is a good starting point and a useful reference for people who are new to the team. It does not update automatically, cannot be read directly by an AI assistant, and creates a version problem the moment anyone improves a prompt and does not republish the PDF.

The version problem compounds over time. After three or four rounds of improvements, the PDF in someone's downloads folder may be a generation behind the version your team actually uses. When that person runs a prompt from the old PDF and gets worse results than expected, they may conclude the prompt does not work rather than that they are using an outdated version.

Treat the PDF as a snapshot: useful for distribution, not for maintenance. Keep the authoritative version somewhere a connected AI can read it, and update the PDF from that source when you need to share a fresh copy with people who prefer a document.

For a guide to building and maintaining a full prompt library, see how to build a prompt library. For an explanation of what a prompt library is and how to organize one, what a prompt library means covers the foundations.

How to do this with fenbs

Use the templates above to create your printable collection, then add the same prompts to fenbs so the live version stays current. Generate a new PDF from fenbs when you need to share a snapshot. Keep the PDF filename or version date so recipients know which generation they have.

1. Create a free account. Add a project for the work your prompt library supports.
2. Go to Skills and prompts in your project. Add each entry from the PDF as its own record: name, prompt text, required inputs, and output check notes.
3. Connect your AI apps at Integrations. A connected assistant can retrieve any prompt from the library before starting a task, without you pasting it manually.
4. When you improve a prompt, update the fenbs entry first, then regenerate the PDF from the current approved text. This keeps the PDF and the live library in sync.